Journal of Reproduction and Fertility (2000) 119, 25–33

Uterine lymphocyte distribution and interleukin expression during early in cows

S. T. Leung1, K. Derecka2,3, G. E. Mann2, A. P. F. Flint2 and D. C. Wathes1*

1Department of Veterinary Basic Sciences, The Royal Veterinary College, Boltons Park, Hawkshead Road, Potters Bar, Hertfordshire EN6 1NB, UK; 2School of Biological Sciences, Division of Animal Physiology, Sutton Bonington Campus, University of Nottingham, Loughborough LE12 5RD, UK; and 3Institute of Animal Reproduction and Food Research of the Polish Academy of Sciences, 10-747 Olsztyn, Poland

Both the production of cytokines and the distribution of immune cells within the change during early pregnancy. Evidence obtained mainly from mice indicates that these changes are important for implantation and in preventing a maternal immune response to the conceptus. The ruminant embryo also produces interferon τ at this time, the signal for the maternal recognition of pregnancy. The relationship between these events in cows was studied using uteri from three groups of animals on day 16 after observed oestrus: (i) cyclic controls, (ii) pregnant and (iii) inseminated but with no embryo present. Embryo size and the antiviral activity in uterine flushings (indicative of the interferon τ concentration) were measured. Sections of intact uterus were frozen for the localization and quantitation of CD4+ (T lymphocytes), CD14+ (macrophages) and CD21+ (B lymphocytes) uterine cells by immunohistochemistry. The expression of interleukin (IL)-1α, IL-2, IL-6 and IL-10 mRNAs in uterine extracts was measured by RT–PCR. Neither embryo size, interferon τ concentration nor pregnancy status influenced the distribution of CD4+, CD14+ or CD21+ cells in the day 16 uterus. Endometrial IL-1α mRNA was detected in most cows across the groups, whereas IL-2 mRNA was only present in the non-pregnant uterus. IL-6 and IL-10 mRNAs were not detectable in any uteri. In conclusion, IL-2 mRNA expression is detectable in the non- pregnant but not the pregnant uterus on day 16 and interferon τ is unlikely to play a role in the redistribution of immune cells in the uterus during early bovine pregnancy.

and delayed-type hormonal responses. The cell-mediated Introduction immune response (Th type 1) accounts for acute cytotoxic An interaction between the conceptus and the immune activity, whereas the hormonal response (Th type 2) involves system is important during implantation (Hansen, 1997). the induction of lymphocyte proliferation and antibody Although the immune system is functional in the uterus production before the rejection of foreign bodies takes place. and the bovine embryo expresses paternal major histo- It is believed that the immune system is biased towards the compatability complex (MHC) molecules of class I on day 7 Th type 2 response during pregnancy (Lin et al., 1993; of pregnancy (Templeton et al., 1987; Low et al., 1990), the Wegmann et al., 1993; Raghupathy, 1997). Type 2 responses conceptus nevertheless escapes the deleterious effect of are associated with preferential production of the cytokines maternal rejection. The exact mechanisms involved in the IL-4, IL-5 and IL-10, whereas in type 1 responses, the pro- maternal tolerance of the foreign fetal antigens during early inflammatory cytokines interleukin 2 (IL-2), interferon γ pregnancy are not fully understood but are likely to involve (IFN-γ) and tumour necrosis factor α (TNF-α) predominate both local production of immunomodulatory signals and (Wegmann et al., 1993; Mosmann and Sad, 1996). Recurrent changes in the type of MHC antigens expressed on the spontaneous has been related to a high degree of trophoblast (Wegmann et al., 1993; Ellis, 1994). It has also maternal immunity during pregnancy (Hill et al., 1992; been suggested that the stimulation of lymphocytes by the Yamada et al., 1994; Hill et al., 1995). In mice, the conceptus may in fact promote conceptus growth by CBA ϫ DBA/2 mating cross has a high rate of spontaneous increasing concentrations of cytokines available at the fetal resorption and their placentae are deficient in IL-4 and conceptus–maternal interface (Wegmann, 1988). IL-10 compared with the more fertile CBA ϫ BALB/c cross Immune responses are mainly divided into cell-mediated (Chaouat et al., 1995). Furthermore, the rate of resorption in the CBA ϫ DBA/2 mice can be reduced by treatment with *Correspondence. either recombinant (r)IL-10 or with r ovine IFN-τ (Chaouat et Received 19 October 1999. al., 1995). Further evidence for the involvement of cytokines

© 2000 Journals of Reproduction and Fertility Ltd 0022–4251/2000 Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access 26 S. T. Leung et al. in fetal loss comes from studies showing increased (0.9% (w/v) NaCl) for collection of embryos. The flushed concentrations of IL-2 in placental tissues associated with uterine fluid was stored frozen at –20ЊC for measurement of spontaneous in both mice (Tangri and Raghupathy, antiviral activity. Complete cross-sections of uterine horns 1993) and women (Hill et al., 1995; Marzi et al., 1996). from the side containing the embryo or corpus luteum were In dairy cows, about 25% of fail in the first 3 frozen in isopentane immersed in liquid nitrogen and stored weeks of gestation, representing a multimillion pound loss to at –70ЊC for immunohistochemistry. Additional samples of the industry (Peters, 1996). In cattle, attachment of the fetal uterine horns were frozen in liquid nitrogen and stored at to the maternal caruncles to form an epitheliochorial –70ЊC for mRNA preparation. placenta does not begin until day 19 of pregnancy (Wathes The length of each flushed blastocyst was measured and and Wooding, 1980). This must be preceded by the signal for the embryos were classified into either small (< 0.5 cm, no the maternal recognition of pregnancy to prevent luteolysis elongation), medium (0.5–1.0 cm, some elongation) or large (Flint et al., 1994a; Wathes and Lamming, 1995). In cows, this (>1.0 cm, clear elongation) size categories. On the basis of signal consists of the production of IFN-τ by the trophoblast the flush results, the cows were classified into three groups: from day 10 to day 25 of pregnancy (Roberts et al., 1992; Flint (i) cyclic controls (n = 8); (ii) pregnant (n = 15); and (iii) et al., 1994b). As well as inhibiting oxytocin receptor inseminated but not pregnant (n = 5). Not all animals were formation and hence luteolysis (Wathes and Lamming, used for all the measurements reported. 1995), IFN-τ regulates lymphocyte proliferation in vitro (Newton et al., 1989; Skopets et al., 1992) and has antiviral activity (Pontzer et al., 1988). Measurement of antiviral activity Studies in ruminants have determined that pregnancy has a profound effect on the uterine population of lymphocytes Interferon activity in uterine flushes was measured by the and macrophages (French and Northey, 1983; Skopets et al., Madin Darby bovine kidney (MDBK) cell Semiliki Forest 1992). The numbers of both types of cell are significantly virus cytoplasmic inhibition assay as described by reduced between early and mid-pregnancy in cattle (Vander Abayasekara et al. (1995). Samples with IFN-τ concentrations Wielen and King, 1984) and sheep (Staples et al., 1983; Lee et initially below the detection limit of 5 antiviral activity units al., 1988; Lee et al., 1992). By mid-pregnancy, virtually no ml–1 were concentrated 10-fold by centrifugal filtration using lymphocytes or macrophages are found in the caruncular MicrocepsTM (Flowgen Instruments Ltd, Lichfield) with a endometrium although they are still present in the inter- molecular cut off of 3000. A starting volume of 3 ml of caruncular endometrium (Gogolin-Ewens et al., 1989; Low et flushed fluid was reconstituted in 300 µl. Samples above the al., 1990). The caruncles are the sites in which fetal chorion detection limit in the initial assay were also concentrated 10- comes into close contact with the maternal endometrium to fold to validate this approach, resulting in a mean Ϯ SEM form placentomes, so these findings indicate that the increase in IFN activity of 913 Ϯ 58%. maternal immune tolerance is both local and specific to the areas adjacent to fetal tissues. Indeed, unknown factors secreted from the ovine placenta also inhibit lymphocyte Immunohistochemistry for localization of uterine activity at various stages of pregnancy (Low et al., 1991). lymphocytes These studies all relate to the period after placental attachment. To date, there has been no investigation of the Immunohistochemistry was carried out on cross-sections uterine lymphocyte population during the critical pre- of uterus as described by Wathes and Hamon (1993) with a implantation period in cattle when high pregnancy losses few modifications. The primary antibodies used were occur. Therefore, the aim of this study was to compare the monoclonal mouse anti-bovine CD4+ (T helper lymphocyte immune cell distribution in the uterus between pregnant and specific, CC8; Naessens et al., 1993), CD14+ (macrophage non-pregnant cows on day 16 using immunohistochemistry. specific, CCG33; Berthon and Hopkins, 1996) and CD21+ (B The results were related to both the embryonic production of lymphocyte specific, CC21; Sopp, 1996) kindly donated by IFN-τ and the concentrations of IL-1α, IL-2, IL-6 and IL-10 Dr Chris Howard (Institute of Animal Health, Compton). mRNAs in the uterine tissue. Serial sections of a bovine lymph node were used as a positive control tissue. Briefly, cross-sections (5 µm thick) were cut on a cryostat and thaw-mounted onto poly-L-lysine- coated (0.1 mg ml–1) slides. The slides were fixed in acetone Materials and Methods (100%) followed by hydrogen peroxide in methanol (1% (v/v); BDH, Poole) at 4ЊC for 30 s each. The slides were Tissue collection washed twice in 1 ϫ PBS for 5 min. All subsequent Holstein–Freisian cows were synchronized with two incubations were carried out in a humidified box at room injections of Estrumate (PGF analogue, Coopers Animal temperature unless otherwise specified. After washing, the Health, Crewe) at 12 days apart and subsequently either sections were incubated with normal rabbit serum (NRS, given a double insemination 72 h and 96 h after the second 2.5%; Sigma, Poole) for 20 min. The NRS was then removed injection or left as uninseminated controls. The animals were by blotting before incubation with 100 µl primary antibody killed on day 16 and the reproductive tract was removed and (collected from the tissue culture supernatant and diluted transported to the laboratory within 30 min. The uterus was 1:5 in 1 ϫ PBS ) at 4ЊC for 1 h. Control sections were treated bisected and each half was flushed with 40 ml sterile saline with non-specific mouse anti-IgG (1 µgml–1; Sigma). An

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access Uterine lymphocytes and interleukin during early pregnancy 27

Table 1. The primer sequences chosen for each individual interleukin

Interleukin Primer Sequence Product (bp) bIL-1α Forward GAC CAC CTC TCT CTC AAT C Reverse CCA TCA CCA CAT TCT CC 146 bIL-2 Forward TTG CAC TAA CTC TGC ACT C Reverse CCAAAA GCAACT GTAAAT CC 108 bIL-6 Forward TAT GAA CTC CCG CTT CAC Reverse CAC ACT CAT CAT TCT TCT CAC 236 oIL-10 Forward AAAACAAGA GCAAGG CGG Reverse CAG GGA CAA GAG AAA TTA GAG G 573 β-actin Forward GAA GAT CTG GCA CCA CAC Reverse AGA GGC GTA CAG GGA CAG C 200 additional CD8+ antibody (specific for cytotoxic T cells) was and β-actin sequence (U39357) (the house keeping gene) tested but did not produce positive staining in control were obtained from Gene Bank. The priming oligonucleotides tissues. Results from this antibody were not included in the used for RT–PCR were subsequently designed using the study. Prime program Wisconsin Package Version 9.0 (Genetics After incubation, the slides were washed twice in 1 ϫ PBS Computer Group (GCG) Madison, WI). The specificity of the for 5 min and incubated with 100 µl per slide of: (i) rabbit anti- identified primers was finally confirmed by the BLAST mouse antiserum (1:100, Sigma) for 30 min; (ii) peroxidase Program (Altschul et al., 1990). The primer sequences chosen anti-peroxidase (1:50, Sigma) for 30 min; and (iii) activated for each individual interleukin are shown (Table 1). diaminobenzidinetetrahydrochloride (DAB; 0.5 mg ml–1; RT–PCR was performed according to a coupled one-step Sigma) for 10 min in the dark. After incubation, the excess procedure (Aatsinki et al., 1994). One microgram of total DAB was removed by washing in distilled water. The slides RNA per 50 µl reaction mixture was reverse transcribed at were then counterstained with haematoxylin and eosin and 42ЊC for 1 h, denatured at 97ЊC for 3 min and amplified for 35 subjected to a serial dehydration in 75%, 100% ethanol, and (IL-1α, IL-10 and β-actin) or 40 (IL-2 and IL-6) cycles. During xylene (100%) each for 2 min. The slides were subsequently each cycle, the reaction mixture was subjected to denaturing mounted for light microscope inspection. at 96ЊC for 30 s, primer annealing at 56ЊC (IL-6, IL-10 and β- Two approximately circular cross-sections of intact uterus actin) or 52ЊC (for IL-1α and IL-2) for 30 s and DNA were processed from each cow with each antibody. The extending at 72ЊC. The enzymes used for the reaction were: stained sections were divided into four quarters from the avian myeloblastosis virus (AMV) reverse transcriptase (4 U; centre and one quarter was used for counting. This quarter Promega, Madison, MI), RNasin (20 U; Promega) and Taq was further sub-divided into three regions: (i) the dense DNA polymerase (1.25 U; Sigma). The PCR products (20 µl) stromal cells immediately underlying the luminal epithelium were subjected to electrophoresis on 1–2% agarose gels and (sub-epithelial stoma); (ii) the remaining endometrium photographed under UV light. RT–PCR control reactions including glands but excluding caruncles; and (iii) the included lymphoid total RNA (positive control) and reaction myometrium. The cells within the luminal epithelium were controls without RNA or reverse transcriptase and bovine not counted as they were few in number. The number of DNA as a template to exclude false positive reactions. Data positively stained cells in each region was counted manually were only included from mRNA samples that generated a by an operator unaware of the pregnancy status of each positive β-actin response. animal. The area covered by each region was measured using The specificity of bIL-1α and bIL-2 RT–PCR products from an image analysis system (Seescan plc, Cambridge). Each uterine RNA were confirmed by Southern blotting. Internal area of measurement covered at least 0.13 mm2. The final oligonucleotides (bIL-1α, 20 mer and bIL-2, 45 mer) labelled results were expressed as the mean Ϯ SEM number of stained with [γ32-P]-ATP using a DNA 5’ end-labelling system cells per mm2 calculated from the two replicate sections from (Promega) were used as probes. The oligonucleotide each animal. sequences used were as follows: bIL-1α: GTG AGG ACC AGA TGA ATA AG; bIL-2: CTC TTG CAC TCG TTG CAA ACG GTG CAC CTA CTT CAA GCT CTA CGG. Pre-hybridization and hybridization were carried out RNA isolation and RT–PCR according to the manufacturer’s instructions. Briefly, the Total RNA from intact bovine uterus was isolated membranes were prehybridized for 4 h at 42ЊC in a solution according to the method of Chomczynski and Sacchi (1987). of 3 ϫ SCC (1 ϫ SCC: 0.015 mol NaCl l–1, 0.015 mol sodium The purity and quality of the RNA extracted was determined citrate l–1, pH 7.0), 50% (v/v) deionized formamide, by the A260:A280 ratio in spectrophotometry and in 5 ϫ Denhardt’s solution, 0.1% (w/v) heat-denatured calf agarose–formaldehyde gel electrophoresis, respectively. thymus DNA and 1% (w/v) SDS. Hybridization was Bovine (bIL) interleukin sequences including bIL-1α performed at 42ЊC overnight after addition of the labelled (accession M37210), bIL-2 (M12791), bIL-6 (X57317) and oligonucleotide probe to the pre-hybridization solution. ovine interleukin (oIL) sequences including oIL-10 (U11421) After hybridization, the membranes were washed twice in

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access 28 S. T. Leung et al.

2 ϫ SSC and 0.1% (w/v) SDS at room temperature. The films distribution among regions was not significantly different. were exposed to X-ray film at –70ЊC for 3–14 days. There was also no significant variation among groups of cows (Table 2). The population of CD14+ macrophages was similar across the three groups in the sub-epithelial stroma. No significant variation in CD14+ cells was found in the Statistical analysis endometrium and myometrium between pregnant and Data were analysed using the SPSS for Windows release control groups (Table 2). CD14+ cells were not counted in the 6.1 statistical program. The relationship between embryo endometrium and myometrium of the inseminated non- size and the antiviral activity in the corresponding flushed pregnant group due to difficulty in identifying the specific fluid was determined using correlation analysis. The regions on the slides from this group. significance of the variation in the number of CD4+, CD14+ Neither the IFN-τ concentration nor the size of the and CD21+ cells, measured as the stained cell count per mm2, embryo was related to the density of CD4+ T cells, CD14+ was determined using a two-way ANOVA (general model) macrophages or CD21+ B cells in the day 16 pregnant uterus. with the pregnancy status of the cows and regions of uterus as factors. Significance was defined as P < 0.05. The expression of interleukins and β-actin in the bovine uterus or lymphoid Expression of interleukin (IL)-1α, IL-2, IL-6 and IL-10 in tissues was identified on the agarose gels as either present the bovine uterus (expressed) or absent (not expressed). No attempt was made to measure the concentration of the RT–PCR product. Bovine lymphoid tissue (positive control) showed expression of mRNAs for bIL-1α, IL-2, IL- 6 and IL-10 as detected by RT–PCR (Fig. 2). For IL-1α mRNA, a single product of 146 bp was expected. Multiple RT–PCR products Results were obtained from both lymphoid and uterine tissues, but only products of about 150 and 110 bp were specific, as Relationship between embryo size and antiviral activity of confirmed by hybridization with an internal oligonucleotide the uterine flushing probe. IL-1α mRNA was present in all the control uteri on Embryos were found in uterine flushings from 15 cows. day 16 of the oestrous cycle (4/4 cows sampled) (Table 3). The concentration of the uterine flushes measured in IL-1α mRNA was also present in 3 of 4 pregnant cows and 2 antiviral units ml–1 (AVU, indicative of the concentration of of 3 cows in the inseminated non-pregnant group on day 16. IFN-τ) associated with the different sizes of embryo were as IL-2 mRNA was present as a single product of 108 bp as expec- follows: small embryos (< 0.5 cm) 4/5 undetectable (< 5 AVU) ted. IL-2 mRNA was not detected in any of the samples of plus 18 AVU, n = 1; medium embryos (0.5–1.0 cm) 125 Ϯ 77 pregnant bovine uterus (0/4), but it was expressed in all but AVU (mean Ϯ SEM), range 5–417 AVU (n = 5); and large one of the non-pregnant cows (6/7 control + inseminated embryos (> 1.0 cm) 215 Ϯ 100 AVU, range 8–593 AVU (n = 5). non-pregnant groups). This finding was confirmed using The antiviral activity was significantly correlated with Southern blotting with an internal oligonucleotide probe for blastocyst length (r2 = 0.33, P = 0.038). IL-2. No products were detected from the RT–PCR reactions with extracts of pregnant uterus. In the lymph tissue, IL-6 was expressed as the expected product of 236 bp and IL-10 as the expected product of 573 bp. Expression of IL-6 and IL-10 Population and distribution of immune cells in the bovine mRNAs was not detected in uterine samples from any of the uterus groups. The lymph node used as the positive control tissue showed intense staining in the germinal centre of the node (in which activated lymphocytes accumulate) for CD4+ Discussion (T lymphocytes) and CD21+ (B lymphocytes); CD14+ (macrophages ) were located mainly in the capsule of the Successful implantation in cows requires an adequate signal lymph node (not shown). for the maternal recognition of pregnancy (Thatcher et al., In the uterus (Fig. 1), the highest concentration of CD4+ 1995; Wathes and Lamming, 1995) and the suppression of a cells was found in the sub-epithelial stroma; much lower hostile maternal immune response (Hansen, 1995, 1997). populations were observed in the remaining endometrium Failure of either of these processes could result in early and myometrium (Table 2). There was a tendency towards embryo loss. IFN-τ is the major signal produced by the higher concentrations of stained cells in the sub-epithelial trophoblast during the preimplantation period. Inadequate stroma of the pregnant cows compared with the control and IFN-τ secretion may be related to pregnancy failure due to inseminated non-pregnant groups but the difference was not the inability of the conceptus to prevent luteolysis (Mann et significant. CD4+ cells were rarely found within the luminal al., 1999). Furthermore, in vitro studies have shown that IFN- epithelium and were therefore not quantified in this location. τ suppresses the activity and proliferation of lymphocytes The CD4+ cells in the endometrium and myometrium caused by trophoblast antigens (Davidson et al., 1994) and usually appeared in a cluster rather than randomly may, therefore, play a role in the onset of maternal immune distributed within the area measured. CD21+ cells also tolerance. The results of the present study confirm that the tended to be highest in the sub-epithelial stroma, but the concentration of IFN-τ, as assessed by antiviral activity, is

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access Uterine lymphocytes and interleukin during early pregnancy 29

(a) (b)

LE LE

SES

Endo

SES

Myo G

(c) (d)

Endo

SES

BV

Myo

Fig. 1. (a) A lower power haematoxylin and eosin stained uterine cross-section to illustrate the histology of the bovine uterus. (b–d) Immunohistochemical staining of CD4+ (T lymphocytes), CD21+ (B lymphocytes) and CD14+ (macrophages) in the bovine uterus. A bovine lymph node was used as the positive control and a negative control, using mouse IgG instead of primary antibody, was also used to validate the staining process (not shown). (b) CD4+ T lymphocytes were found mainly in the cells of the sub-epithelial stroma. (c) CD21+ B lymphocytes in the myometrium, where they were often located around the blood vessels. (d) CD14+ macrophages were also present in the sub-epithelial stroma. The sections illustrated are from the pregnant group (b,d) and the cyclic control group (c), although no differences in distribution were detected between pregnant and non-pregnant cows (see Table 2). BV: blood vessels; Endo: endometrium; G: glandular epithelium; LE: luminal epithelium; Myo: myometrium; SES: sub-epithelial stroma. Scale bars represent (a) 750 µm, (b) 75 µm, (c) 250 µm, (d) 75 µm.

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access 30 S. T. Leung et al.

Table 2. Distribution of CD4+, CD14+ and CD21+ cells in the bovine uterus on day 16 of the oestrous cycle or early pregnancy

CD4+ CD21+ CD14+ Area Samplea Number of cows (T lymphocytes) (B lymphocytes) (Macrophages)

Sub-epithelial stroma Preg 7–9 239 Ϯ 51.0 151 ± 49.3 527 ± 86.3 Cont 6–8 214 Ϯ 26.7 210 Ϯ 50.7 487 Ϯ 129.0 INP 4–5 156 Ϯ 85.9 131 Ϯ 45.0 383 Ϯ 106.4 Endometrium Preg 5 44 Ϯ 11.3 44 Ϯ 15.5 435 Ϯ 105.2 Cont 5 53.5 Ϯ 21.1 87 Ϯ 16.9 320 Ϯ 130.5 INP 5 59 Ϯ 21.1 63 Ϯ 11.2 ND Myometrium Preg 5 30 Ϯ 7.0 59 Ϯ 7.0 78 Ϯ 12.0 Cont 5 14 Ϯ 8.0 107 Ϯ 21.0 187 Ϯ 97.4 INP 5 30 Ϯ 12.7 87 Ϯ 8.4 ND

The number of stained cells per mm2 was measured in two replicate sections per cow. Values are mean ± SEM. No significant variation was found in the number of cells within a region among any of the groups. aAnimals were classified as either uninseminated controls (Cont), pregnant (Preg) or inseminated but not pregnant (INP). ND: not done.

IL-1a IL-2 IL-6 IL-10 bp Lu L uLu bp Lu

700 200

600

500 100

400

Fig. 2. RT–PCR products of bovine interleukin (bIL)-1α, IL-2, IL-6 and IL-10 mRNA. Products of bIL-1α, IL-2 and IL-6 were subjected to electrophoresis in 2% agarose gels (whereas bIL-10 was in a 1% agarose gel) and photographed under ultraviolet (UV) light. The size of the molecular weight markers is given on the left in base pairs (bp). Arrows show the expected size of the RT–PCR products. The identity of the bIL-1α and IL-2 products were confirmed by Southern blotting using internal oligonucleotides. L: lymphoid (positive control); u: uterus. proportional to the size of the embryo. Although the a relatively high population of lymphocytes in the sub- embryos were at the same stage of pregnancy, the largest epithelial stroma in the ovine uterus throughout the oestrous embryo produced more than 10-fold more IFN-τ than the cycle and showed that the number of intra-epithelial smallest embryo. It is possible that cows with smaller lymphocytes was only reduced between day 18 and day 46 of embryos would be less likely to maintain their pregnancies, pregnancy. This lack of change before implantation is in although this could not be determined from the present agreement with the results of the present study. These study, which terminated at day 16. observations indicate that the immunosuppressive However, in the present study, neither the concentration of mechanism affecting lymphocyte proliferation during IFN-τ nor the size of the embryo was correlated with the pregnancy was not effective until after the first 3 weeks of density of CD4+ T cells, CD14+ macrophages or CD21+ B cells gestation. This finding implies that IFN-τ does not have a in the day 16 pregnant uterus. T (CD4+) and B (CD21+) direct effect on lymphocyte redistribution, as IFN-τ lymphocytes were rarely found within either the cyclic or production ceases after implantation whereas maternal pregnant luminal epithelium or in the uterine lumen. Intra- immune tolerance persists for the rest of the pregnancy. epithelial CD8+ lymphocytes have been identified in sheep IL-2 is a pro-inflammatory cytokine. A novel finding of the (Hansen, 1995), although in the present study it was not present study was that IL-2 mRNA was not detectable in the possible to stain for this type of cell with the antibodies pregnant bovine uterus whereas it was present in the non- available. Thus, possible changes in the epithelial pregnant cows on day 16. It has been demonstrated that lymphocyte population might not have been detected. PGE2 inhibits the expression of IL-2 in bovine lymphocytes

However, Vander Wielen and King (1984) identified (Emond et al., 1998) and that the basal synthesis of PGE2 in lymphocytes in the bovine luminal epithelium and sub- the endometrium is stimulated by IFN-τ (Danet-Desonyers et epithelial stroma and found no significant variation in the al., 1994). Similarly, Bergeron et al. (1996) reported that day 12 number of lymphocytes in these regions between cyclic and blastocoelic fluid from rabbits inhibited IL-2 gene expression pregnant animals on day 20. Lee et al. (1988, 1992) also found in stimulated lymphocytes. This finding provides a

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access Uterine lymphocytes and interleukin during early pregnancy 31

Table 3. mRNA expression of interleukin (IL)-1α, IL-2, IL-6 al., 1992). In the pseudopregnant mouse uterus, peaks of IL-6 and IL-10 in the bovine uterus on day 16 of the oestrous cycle bioactivity were detected between day 4 and day 8 or early pregnancy (Choudhuri and Wood, 1993). Although IL-6 mRNA and protein were not detected in fresh uterus, they increased a β α Sample -actin IL-1 IL-2 IL-6 IL-10 after culture of uterine epithelial and stromal cells (Jacobs et al., 1992; Kover et al., 1995). IL-6 mRNA has also been Preg 4/4 3/4 0/4 0/4 0/4 Cont 4/4 4/4 3/4 0/4 0/4 reported in elongating bovine, ovine and porcine INP 3/3 2/3 3/3 0/3 0/3 conceptuses (Mathialagan et al., 1992), although the protein Lymph node + + + + + inhibits embryonic growth and attachment in vitro in mice (Jacobs et al., 1992). IL-6 mRNA was not detected in any Interleukin expression was classified as either present or absent (not uterine samples in the present study, indicating that it is not expressed) on the basis of a correctly sized band. No attempt was made to an important regulator of uterine activity at this stage of measure the concentration of the RT–PCR product. β-Actin was used as a positive control for RT–PCR and a bovine lymph node was used as a positive pregnancy in cows. control for the detection of the relevant interleukins. IL-10 is a Th type 2 anti-inflammatory cytokine which aAnimals were classified as either uninseminated controls (Cont), pregnant inhibits the production of cytokines that stimulate Th1 (Preg) or inseminated but not pregnant (INP). immune responses (Moore et al., 1993). IL-10 is produced by +, positive RT–PCR result. human placenta during gestation and the concentration decreases only at parturition (Paradowska et al., 1997; mechanism to explain the difference in IL-2 expression and Simpson et al., 1998). IL-10 is also an important placental indicates another potentially important role for IFN-τ during product in mice and anti-IL-10 can increase the fetal early pregnancy. Comparative information on uterine IL-2 resorption rate in CBA ϫ DBA/2 matings (Chaouat et al., expression is limited. Jokhi et al. (1997) were unable to detect 1995). IL-10 mRNA was not detected in the pregnant bovine IL-2 protein or mRNA in first trimester human endometrium uterus on day 16 in the present study. This finding was and Fox et al. (1998) did not detect IL-2 mRNA in uterine unexpected, but it is possible that IL-10 may be produced by intercaruncular epithelium in late pregnant ewes. In either fetal or placental tissue later in pregnancy. contrast, Grunig and Antczak (1995) detected IL-2 mRNA in In conclusion, the lack of significant variation of CD4+ T horse endometrium but not trophoblast between day 30 and cells, CD21+ B cells and CD14+ macrophages in the bovine day 55 of gestation, but mares are unusual in that they mount uterus on day 16 between pregnant and non-pregnant cows a strong immune response to invading trophoblast. indicates that differentiation of the maternal immune IL-1α is also a pro-inflammatory cytokine. The IL-1 system response from Th1 to Th2 may occur at a later stage of has two ligands, IL-1α and IL-1β, which share 26% sequence pregnancy and that IFN-τ does not have a role in this homology and act through the same receptor. In the present process. The pattern of uterine cytokine expression on day 16 study, IL-1α was present in most uteri from both cyclic and suggests that IL-1 production in the late luteal phase, acting τ, pregnant cows on day 16. This is consistent with previous in concert with IFN- stimulates stromal PGE2 synthesis and reports: in the mouse uterus, both IL-1 mRNA and that this in turn inhibits uterine IL-2 expression. This effect bioactivity increase between day 4 and day 8 of may help to prevent a maternal immune response to pseudopregnancy associated with the process of implantation. decidualization (Choudhui and Wood, 1993) and IL-1α and IL-1β mRNA expression in isolated uterine epithelial cells This work was supported by the BBSRC, the Ministry of increases during a 4 day culture period (Kover et al., 1995). Agriculture, Fisheries and Food and the Milk Development Council. Similarly, IL-1α mRNA was present in sheep endometrial The bovine monoclonal antibodies were kindly donated by Chris epithelium (Fox et al., 1998), and in women, IL-1β Howard, Institute of Animal Health, Compton, UK. The authors thank Sarah Mann and Pat Fisher for technical assistance and immunoreactivity in the endometrium increased in the late Barbara Wilsmore for photography. secretory phase (Tabibzadeh, 1991). IL-1β was present in uterine flushes from both pregnant and non-pregnant cows on days 14 and 17, whereas it was not detectable on days 25 References and 30 (Davidson et al., 1995). IL-1 stimulates prostaglandin synthesis, particularly PGE through a mechanism involving Aatsinki JT, Pietilä EM, Lakkakorpi JT and Rajaniemi HJ (1994) A coupled 2 one-step reverse transcription PCR procedure for generation of full-length increased cyclo-oxygenase (Bany and Kennedy, 1995; open reading frames Biotechniques 16 282–288 Davidson et al., 1995). IL-1α also plays an important role in Abayasekara DRE, Searle MJ and Flint APF (1995) Interferon tau activates modifying the endometrial structure for implantation and protein kinase C in Madin Darby bovine kidney cells Biochemical Society during the human menstrual cycle (Singer et al., 1997; Huang Transactions 23 271S et al., 1998). An IL-1 type I receptor antagonist administered Altschul S, Gish W, Miller W, Myers E and Lipman D (1990) Basic local alignment search tool Journal of Molecular Biology 21 5403–5410 from days 3–9 of pregnancy causes an almost complete block Bany BM and Kennedy TG (1995) Interleukin-1α regulates prostaglandin of implantation in mice (Simon et al., 1994). Together, these production and cyclo-oxygenase activity in sensitized rat endometrial results indicate that uterine production of IL-1 plays an stromal cells in vitro. Biology of Reproduction 53 126–132 important role during implantation in a variety of species. Bergeron D, Audette M and Lambert RD (1996) Regulation of leukocyte interleukin 2 and interleukin 2 receptor gene expression by rabbit In the human uterus, IL-6 immuno- and bioactivity have blastocoelic fluid Journal of Reproduction and Fertility 106 143–151 been reported in both stromal and epithelial cells throughout Berthon P and Hopkins J (1996) Ruminant cluster CD14 Veterinary the cycle (Tabibzadeh et al., 1989; Tabibzadeh, 1991; Jacobs et Immunology and Immunopathology 52 245–248

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access 32 S. T. Leung et al.

Chaouat G, Melian AA, Martal J, Raghupathy R, Elliot J, Mosmann T and uterus of non-pregnant and pregnant sheep American Journal of Reproductive Wegmann TG (1995) IL-10 prevents naturally occurring fetal loss in the Immunology 28 90–96 CBAx DBA/2 mating combination, and local defects in IL-10 production in Lin H, Mosmann TR, Guilbert L, Tuntipopipat S and Wegmann TG (1993) this -prone combination is corrected by in vivo injection of IFN-τ Synthesis of T helper 2-type cytokines at the maternal–fetal interface Journal Journal of Immunology 154 4261–4268 of Immunology 151 4562– 4573 Chomczynski P and Sacchi N (1987) Single-step method RNA isolation by Low BG and Hansen PJ (1998) Actions of steroids and prostaglandins secreted acid guanidinium thiocyanate–phenol–chloroform extraction Analytical by the placenta and uterus of the cow and ewe on lymphocyte proliferation in Biochemistry 162 156–159 vitro. American Journal of Reproductive Immunology and Microbiology 18 71–75 Choudhuri R and Wood GW (1993) Production of interleukin-1 and Low BG, Hansen PJ and Drost M (1990) Expression of major interleukin-6 and tumor necrosis factor alpha in the uterus of histocompatibility complex antigens on the bovine placenta Journal of pseudopregnant mice Biology of Reproduction 49 596– 603 Reproduction and Fertility 90 235–243 Dannet-Desnoyers G, Wetzels C and Thatcher WW (1994) Natural and Low BG, Hansen PJ and Drost M (1991) Inhibition of in vitro lymphocyte recombinant bovine interferon tau regulate basal and oxytocin induced proliferation by ovine placenta-conditioned culture medium Journal of

secretion of prostaglandin F2α and E2 by epithelial cells and stromal cells in Reproductive Immunology 19 25–41 the endometrium Reproduction, Fertility and Development 6 193–202 Mann GE, Lamming GE, Robinson RS and Wathes DC (1999) The regulation Davidson JA, Betts JG, Tiemann U, Kamwanja LA, Monterroso VH and of interferon τ production and uterine hormone receptors during early Hansen PJ (1994) Effects of interferon-tau and interferon-alpha on pregnancy Journal of Reproduction and Fertility Supplement 54 317–328 proliferation of bovine endometrial cells Biology of Reproduction 51 700–705 Marzi M, Vigano A, Trabattoni D, Villa ML, Salvaggio A, Clerici E and Davidson JA, Tiemann U, Betts JG and Hansen PJ (1995) DNA synthesis and Clerici M (1996) Characterization of type 1 and type 2 cytokine production prostaglandin secretion by bovine endometrial cells is regulated by IL-1 profile in physiologic and pathologic human pregnancy Clinical and Reproduction, Fertility and Development 7 1037–1043 Experimental Immunology 106 127–133 Ellis SA (1994) MHC studies in domestic animals European Journal of Mathialagan N, Bixby JA and Roberts RM (1992) Expression of interleukin-6 Immunogenetics 21 209–215 in porcine, ovine and bovine preimplantation conceptuses Molecular

Emond W, Fortier MA, Murphy BD and Lambert RD (1998) Prostaglandin E2 Reproduction and Development 32 324– 330 regulates interleukin-2 and granulocyte–macrophage colony-stimulating Moore KW, O’Garra A, Malefyt RDW, Vieira P and Mosmann TR (1993) factor gene expression in bovine lymphocytes Biology of Reproduction 58 Interleukin-10 Annual Review of Immunology 11 165–190 143–151 Mosmann TR and Sad S (1996) The expanding universe of T-cell subsets: Th1, Flint APF, Lamming GE, Stewart HJ and Abayasekara DRE (1994a) The role Th2 and more Immunology Today 17 138–146 of the endometrial oxytocin receptor in determining the length of the sterile Naessens J, Olubayo R, Davis NC and Hopkins J (1993) Cross-reactivity of oestrous cycle and ensuring maintenance of luteal function in early workshop antibodies with cells from domestic and wild ruminants pregnancy in ruminants Philosophical Transaction of the Royal Society of Veterinary Immunology and Immunopathology 39 283–290 London Series B – Biological Science 344 291–304 Newton GR, Vallet JL, Hansen PJ and Bazer FW (1989) Inhibition of Flint APF, Guesdon FMJ and Stewart HJ (1994b) Regulation of trophoblast lymphocyte proliferation by ovine trophoblast-1 and a high molecular interferon gene Molecular and Cellular Endocrinology 146 108–113 weight glycoprotein produced by the peri-implantation sheep conceptus Fox A, Lee CS, Brandon MR and Meeusen EN (1998) Effects of pregnancy on American Journal of Reproductive Immunology 19 99–107 lymphocytes within sheep interplacentomal epithelium American Journal of Paradowska E, Blach-Olszewska Z and Gejdel E (1997) Constitutive and Reproductive Immunology 40 295–302 induced cytokine production by human placenta and amniotic membrane French LR and Northey DL (1983) Inhibitory effect of the bovine conceptus on at term Placenta 18 441–446 lymphocyte stimulation Journal of Animal Science 57 456–465 Peters AR (1996) Embryo mortality in the cow Animal Breed Abstract 64 587–598 Gogolin-Ewens KJ, Lee CS, Mercer WR and Brandon MR (1989) Site-directed Pontzer CH, Torres BA, Vallet JL, Bazer FW and Johnson HM (1988) Antiviral differences in the immune response to the Immunology 66 312–317 activity of the pregnancy recognition hormone ovine trophoblast protein-1 Grunig G and Antczak DF (1995) Horse trophoblasts produce tumor necrosis Biochemical and Biophysical Research Communications 152 801–807 factor α but not interleukin 2, interleukin 4 or interferon γ Biology of Raghupathy R (1997) Th1-type immunity is incompatible with successful Reproduction 52 531–539 pregnancy Immunology Today 18 478–482 Hansen PJ (1995) Interactions between the immune system and the ruminant Roberts RM, Leaman WW and Cross JC (1992) Role of interferons in maternal conceptus Journal of Reproduction and Fertility Supplement 49 69–82 recognition of pregnancy in ruminants Proceedings of the Society of Experimental Hansen PJ (1997) Interactions between the immune system and the bovine Biology and Medicine 200 7–18 conceptus Theriogenology 47 121–130 Simon C, Francis A and Piquette G (1994) Embryonic implantation in mice is Hill JA, Polgar K and Harlow BL (1992) Evidence of embryo- and trophoblast- blocked by interleukin-1 receptor antagonist Endocrinology 134 521–528 toxic cellular immune response(s) in women with recurrent spontaneous Simpson KL, Keelan JA and Mitchell MD (1998) Labor-associated changes in abortion American Journal of Obstetrics and Gynecology 166 1044–1052 interleukin-10 production and its regulation by immunomodulators in human Hill JA, Polgar K and Anderson DJ (1995) T-helper 1-type immunity to choriodecidua Journal of Clinical Endocrinology and Metabolism 83 4332–4337 trophoblast in women with recurrent spontaneous abortion JAMA 273 Singer CF, Marbaix E, Kokorine I, Lemoine P, Donnes J, Eeckhout Y and 1933–1936 Courtoy PJ (1997) Paracrine stimulation of interstitial collagenase (MMP-1) Huang HY, Wen Y, Irwin JC, Kruessel JS, Soong YK and Polan ML (1998) in the human endometrium by interleukin 1α and its dual block by ovarian Cytokine-mediated regulation of 92-kilodalton type IV collagenase, tissue steroids Proceedings of the Society of Experimental Biology and Medicine 94 10 inhibitor of metalloproteinase-1 (TIMP-1), and TIMP-3 messenger 341–10 345 ribonucleic acid expression in human endometrial stromal cells Journal of Skopets B, Li J, Thatcher WW, Roberts RM and Hansen PJ (1992) Inhibition of Clinical Endocrinology and Metabolism 83 1721–1729 lymphocyte proliferation by bovine trophoblast protein-1 (type 1 trophoblast Jacobs AL, Sehgal PB, Julian J and Carson DD (1992) Secretion and hormonal interferon) and bovine interferon-alpha 1 Veterinary Immunopathology 34 81–96 regulation of interleukin-6 production by mouse uterine stromal and Sopp P (1996) Ruminant cluster CD21 Veterinary Immunology and polarised epithelial cells cultured in vitro. Endocrinology 131 1037–1046 Immunopathology 52 249 Jokhi PP, King A and Loke YW (1997) Cytokine production and cytokine Staples LD, Heap RB, Wooding FBP and King GJ (1983) Migration of receptor expression by cells of the human first trimester placental–uterine leucocytes into the uterus after acute removal of ovarian progesterone interface Cytokine 9 126–137 during early pregnancy in the sheep Placenta 4 339–349 Kover K, Laing L, Andrews GK and Dey SK (1995) Differential expression Tabibzadeh S (1991) Human endometrium: an active site of cytokine and regulation of cytokine genes in the mouse uterus Endocrinology 136 production and action Endocrine Reviews 12 272–290 1666–1673 Tabibzadeh S, Santhanam U, Sehgal PB and May LT (1989) Cytokine induced Lee CS, Meeusen E, Gogolin-Ewens K and Brandon MR (1988) Identification production of IFN-β/IL-6 by freshly explanted human endometrial stromal of a unique lymphocyte subpopulation in the sheep uterus Immunology 63 cells Journal of Immunology 142 3134–3139 157–164 Tangri S and Raghupathy R (1993) Expression of cytokines in mice Lee CS, Meeusen E, Gogolin-Ewens K and Brandon MR (1992) Quantitative undergoing immunologically mediated spontaneous fetal resorptions and qualitative changes in the intraepithelial lymphocyte population in the Biology of Reproduction 49 850–865

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access Uterine lymphocytes and interleukin during early pregnancy 33

Templeton JW, Tipon RC, Garber T, Bondioli K and Kraemer DC (1987) maintenance of pregnancy Journal of Reproduction and Fertility Supplement 49 Expression and genetic segregation of parental BoLA serotype in bovine 53–67 embryo Animal Genetics 18 317– 322 Wathes DC and Wooding FBP (1980) An electron microscopic study of Thatcher WW, Meyer MD and Danet-Desonoyers G (1995) Maternal implantation in the cow American Journal of Anatomy 159 285–306 recognition of pregnancy Journal of Reproduction and Fertility Supplement 49 Wegmann TG (1998) Maternal T cells promote placental growth and prevent 15–28 spontaneous abortion Immunology Letters 17 297–302 Vander Wielen AL and King GJ (1984) Intraepithelial lymphocytes in the Wegmann TG, Lin H, Guilbert L and Mosmann TR (1993) Bidirectional bovine uterus during the oestrous cycle and early gestation Journal of cytokine interactions in the maternal–fetal relationship: is successful Reproduction and Fertility 70 457–462 pregnancy a Th2 phenomenon? Immunology Today 14 353–356 Wathes DC and Hamon M (1993) Localisation of oestradiol, progesterone and Yamada H, Polgar K and Hill JA (1994) Cell-mediated immunity to oxytocin receptors in the uterus during the oestrous cycle and early trophoblast antigens in women with recurrent spontaneous abortion pregnancy of the ewe Journal of Endocrinology 138 479–491 American Journal of Obstetrics and Gynecology 170 1339–1344 Wathes DC and Lamming GE (1995) The oxytocin receptor, luteolysis and the

Downloaded from Bioscientifica.com at 10/02/2021 02:51:29AM via free access