- Email: [email protected]
SCID mice
Journal of Reproductive Immunology 68 (2005) 39–51
Lymphocyte phenotyping and NK cell activity analysis in pregnant NOD/SCID mice Yi Lin ∗ , Yijing Chen, Yaoying Zeng, Tong Wang, Shan Zeng Institute of Tissue Transplantation and Immunology, Jinan University, Guangzhou City 510632, China Received 1 April 2005; received in revised form 9 May 2005; accepted 9 May 2005
Abstract The purpose of this study was to outline the fertility features of non-obese diabetic (NOD)/LtSzscid/scid (NOD/SCID [severe combined immunodeficiency] for short) mice, and to evaluate the effects of NK cell subsets on the pregnancy outcomes of the syngeneic NOD/SCID × NOD/SCID mating combination. Firstly, lymphocyte phenotyping was performed with flow cytometry to confirm the multiple immunodeficits in NOD/SCID mice. Fertility features were assessed in NOD/SCID × NOD/SCID mice and compared with non-immunodeficiency control BALB/c × BALB/c mice. Although the presence of NK cell deficit is apparent in NOD/SCID mice, a certain level of remnant NK activity could be observed in these mice. The remnant NK cell activity was stimulated with polyinosinic-polycytidylic acid (polyIC), or inhibited with anti-asialo GM1 (ASGM1) anti-serum, respectively. The effects of these factors on pregnancy outcomes were evaluated after administration. Roughly normal fertility could be observed in NOD/SCID × NOD/SCID mice. However, a slightly larger sized litter was observed in polyIC-treated NOD/SCID × NOD/SCID mice compared with control NOD/SCID mice. In contrast, embryo resorption was boosted after ASGM1 injection, and correlated subsequently with a smaller litter size. It indicates that the remnant NK cell activity in NOD/SCID mice may be beneficial to feto-maternal tolerance during pregnancy. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Animal model; Immunodeficiency diseases; Placenta; Reproductive immunology
Abbreviations: ASGM1, anti-asialo GM1; FSC, forward scatter count; mAb, monoclonal antibody; NOD, non-obese diabetic mouse; polyIC, polyinosinic-polycytidylic acid; SCID, severe combined immunodeficiency; SSC, side scatter count; uNKcell, uterine NK cell ∗ Corresponding author. Tel.: +86 20 85220298; fax: +86 20 85221337. E-mail address: [email protected] (Y. Lin). 0165-0378/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.jri.2005.05.002
40
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
1. Introduction Bosma et al. (1983) originally reported that C.B-17 mice homozygous for the severe combined immunodeficiency (scid/scid) mutation lack functional T and B lymphocytes. Although they are defective in cellular and humoral immune function, these mice display normal NK cell function (Dorshkind et al., 1985). On the other hand, the non-obese diabetes (NOD)/Lt strain is characterized by a functional deficit in natural killer (NK) cells (Kataoka et al., 1983) and tends to develop a high incidence of autoimmune insulin-dependent diabetes mellitus (IDDM; Christianson et al., 1993). Shultz et al. (1995) have established a murine model, NOD/LtSz-scid/scid (NOD/SCID [severe combined immunodeficiency] for short), characterized by multiple defects in innate immunity as well as in the absence of T and B cell function, by backcrossing of the scid mutation onto the NOD/Lt strain background. Although NK cell deficit is apparent in NOD/SCID mice, remnant NK cell activity could be detected in these mice (Christianson et al., 1997). To our knowledge, little is known about the remnant function of these cells in feto-maternal tolerance during pregnancy. In addition, as a result of T cell deficiency, NOD/SCID mice do not develop IDDM, which is mediated by T cells (Shultz et al., 1995). As early as 1991, King and Loke (1991) (Loke and King, 1991) had spotted that uterine NK (uNK) cells “may have a role in the control of implantation and the transformation of the uterine vasculature by trophoblast on which the blood supply to the feto-placental unit depends,” at a time when NK cells were only seen as candidate abortificient cells affecting mice. However, in CBA/J × DBA/2 mice, a widely used model of spontaneous abortion, activated NK cells and macrophages, was believed to be associated with the increase in embryo resorption (Clark et al., 1998). In the present study, T cell, B cell and NK cell subsets were analyzed with flow cytometry (FCM) to investigate whether there are any deficits in NOD/SCID mice. The fertility features were evaluated in these mice, and compared with a non-immunodeficiency model of BALB/c × BALB/c. The remnant NK cell activity in NOD/SCID mice was investigated and stimulated with polyinosinic-polycytidylic acid (polyIC), an activator of NK cells. In another design, inhibition of NK cells was performed by an i.p. injection of anti-asialo GM1 (ASGM1) anti-serum at an earlier stage of pregnancy in these mice. To our knowledge, CD94 is expressed mainly by NK cells and is present as a heterodimer with NKG2 on the cell surface. It plays an important role in the adhesion and activation of NK cell lineage (Vance et al., 2002). Therefore, the DX5+ CD94+ /DX5+ cell percentage in placental NK cells was determined with FCM using DX5 (its monoclonal antibody [mAb] is specific for CD49b) as a pan-NK cell marker (Arase et al., 2001). The effects of NK cells on pregnancy outcomes were evaluated in NOD/SCID × NOD/SCID mice.
2. Materials and methods 2.1. Mice BALB/c and NOD/LtSz-scid/scid mice (NOD/SCID for short) were purchased from the Experimental Animal Center of Zhongshan University (Guangzhou City, China), and bred
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
41
Table 1 Flow cytometric analysis of CD3+ , CD19+ and DX5+ cells with CD45 as a leukocyte common marker (n = 6 for each group) Cell source
Cell subset
Cell percentage
BALB/c
NOD/SCID
Peripheral blood Peripheral blood Placenta
T cell B cell NK cell
CD45+ CD3+ /CD45+ CD45+ CD19+ /CD45+ CD45+ DX5+ /CD45+
72.3 ± 6.8 3.6 ± 0.3 72.5 ± 3.2
4.3 ± 0.6a 1.3 ± 0.2a 6.4 ± 0.5a
NK, natural killer; NOD, non-obese diabetic mice; SCID, severe combined immunodeficiency. a Independent sample’s t-test: P < 0.001 vs. BALB/c group.
under specific pathogen-free conditions in the laboratory for immunodeficiency animals (Jinan University, Guangzhou, China). Each mouse was considered sexually mature, based on an age of 10–12 weeks. A syngeneic NOD/SCID × NOD/SCID breeding scheme was established, using the BALB/c × BALB/c mating combination as a non-immunodeficiency control. Lymphocyte phenotyping was performed in virgin mice and pregnant mice on day 13.5 of gestation. Glucosuria was assessed weekly in NOD/SCID mice for 4 months using Test strips (Baiyuan Mountain Pharmaceutical, Guangzhou, China). H&E-stained sections of thymus were examined to assess whether thymic lymphoma had developed in these mice. As expected, no urinary glucose was found in these mice, confirming their resistance to diabetes (Shultz et al., 1995). In addition, no thymic lymphoma was found in these mice at ages ranging from 12 to 30 weeks (Christianson et al., 1993; Shultz et al., 1995). Based on methods published previously (Shultz et al., 1995), we confirmed the T cell and B cell deficits in the NOD/SCID with FCM, using CD45, CD3 and CD19 as leukocyte common antigen, T cell marker and B cell marker, respectively. CD3+ cells in peripheral blood mononuclear cells (PBMC) were measured at 5.9% of the level found in control BALB/c mice, while CD19+ cells were 36.1% of the level in BALB/c, supporting the combined deficits in T cell and B cell subsets in the NOD/SCID (Table 1, P < 0.001). NK cell deficit in these mice was confirmed using placental cells and methods described in Sections 2.4 and 2.6. 2.2. Fertility features of NOD/SCID × NOD/SCID mice In the hope that pregnancy would result, one female was co-caged with one male, and on day 0.5 a vaginal plug was observed. Animals were housed under barrier conditions and monitored for health status. In addition, they were subjected to daily 12-h alternating periods of light and dark cycles within a controlled environment of 63% humidity at a temperature of 72–74 F. Food and water were provided ad libitum. Female mice were weighed twice a week. Data on each breeding pair were collected over a 4month period beginning with the establishment of the breeding (Tempfer et al., 2000). In another design, pregnant mice were sacrificed on day 13.5 of gestation. Peripheral blood, spleens, placentas and thymus were harvested. Collection of peripheral blood and placentas and purification of lymphocytes were performed using methods described in Section 2.6. The resorption rate of the embryos was calculated as follows: resorption
42
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
rate (%) = [number of resorbed embryos/number of total (resorbed + viable) embryos] × 100. 2.3. In vivo stimulation of NK cells Remnant NK cell activity was boosted by polyIC administration. Pregnant NOD/SCID mice were injected peritoneally with polyIC (␥-irradiated, Sigma Chemical, St. Louis, MO, USA), at a dosage of 100 g/20 g body weight (Shimada et al., 2003), while pregnant BALB/c and NOD/SCID mice injected with phosphate buffer solution (PBS) were used as controls. The injection was given on days 4.5, 7.5 and 10.5 of gestation consecutively. The efficacy of such treatment was evaluated by FCM using NK cell markers and YAC-1 cellular cytotoxicity assays, as described as follows (Lee et al., 2004; Shultz et al., 1995). 2.4. Assessment of NK cell activity The NK cell activity was determined as previously described (Lee et al., 2004; Shultz et al., 1995). Briefly, splenocytes were harvested from polyIC- and ASGM1-treated mice on day 13.5 of gestation, respectively. The target YAC-1 cells (American Type Culture Collection, Rockville, MD, USA) were labeled with 51 Cr and adjusted to 5 × 104 cells/ml. Various E:T ratios were set up in triplicate in V-bottom plates. Supernatants were harvested for counting on a gamma counter 4 h after incubation of targets and effectors at 37 ◦ C. The percentage-specific 51 Cr release was calculated using the following formula, where X is the mean experimental release from triplicate wells, total release (T) was determined from wells receiving 51 Cr-labeled YAC-1 target cells and 2% SDS, and spontaneous release (S) was determined from wells receiving 51 Cr-labeled YAC-1 target cells in growth medium: % specific release = [(X − S)/(T − S)] × 100. 2.5. In vivo inhibition of NK cells Inhibition of NK cells was performed using ASGM1 anti-serum (Wako Pure Chemical Inc., Osaka, Japan). NOD/SCID mice were injected peritoneally with 50 l of ASGM1 or an equivalent amount of PBS on days 4.5, 7.5 and 10.5 of gestation consecutively. The efficacy of such treatment was evaluated by FCM using NK cell markers and using standard YAC-1 cellular cytotoxicity assays, as mentioned above. 2.6. Monoclonal antibodies and flow cytometry The following mAbs were purchased from eBioscience (San Diego, CA, USA): fluorescein isothiocyanate (FITC)-conjugated CD3 (clone: 145-2C11); FITC- and phycoerythrin (PE)-conjugated CD45 (30-F11); FITC- and PE-conjugated DX5 (CD49b, clone: DX5); FITC-conjugated CD19 (MB19-1) and PE-conjugated CD94 (18d3). As previously described (Lin et al., 2004; Shultz et al., 1995), mononuclear cells were isolated from peripheral blood in virgin mice and peripheral blood and placentas in pregnant mice on day 13.5 of gestation. Briefly, peripheral blood was harvested from vena orbitalis, heparinized and purified by centrifugation on ficoll-hypaque density medium and incubated
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
43
with red cell lysis buffer. The isolated PBMCs were used for the confirmation of T cell and B cell deficits as mentioned above in Section 2.1. Hysterolaparotomy was performed to collect embryo-depleted placentas and associated decidual tissue, including decidua basalis. The pooled placentas were carefully cut into small pieces less than 1 mm3 with ocular scissors, and the pieces were then collected in Hanks’ balanced salt solution (HBSS) and filtered through a 50-m pore size nylon mesh to obtain mononuclear cell suspension. Mononuclear cells were purified by centrifugation of cell suspension on ficoll-hypaque density medium. Similarly, any contaminating red blood cells that might have persisted in the single-cell suspension were eliminated by incubation with red cell lysis buffer. Cells were incubated in the indicated mAb conjugates for 30 min in a total volume of 50 l of PBS containing 3% bovine serum albumen (BSA). After being washed twice with PBS, these cells were resuspended in 1% paraformaldehyde. Immunostained cells were analyzed on a FACSCalibur flow cytometer using CellQuest software (Becton Dickinson, San Jose, CA, USA). Ten thousand cells were detected in each sample. For a negative control, adequately designed gate thresholds were set to capture lymphoid cells using appropriate FITC- and PE-conjugated isotype control antibodies (eBioscience). 2.7. Statistical analysis A Student’s t-test was used to determine statistical significance where necessary. The resorption rate of embryos was analyzed using the χ2 -test.
3. Results 3.1. Confirmation of T, B and NK cell deficits in NOD/SCID mice In the current study, T cell and B cell deficits were confirmed by comparing the CD45+ CD3+ /CD45+ and CD45+ CD19+ /CD45+ cell percentages in PBMCs in nonpregnant NOD/SCID mice and BALB/c mice, respectively. NK cell deficit was confirmed by comparing the CD45+ DX5+ /CD45+ cell percentage in placental lymphocytes in NOD/SCID mice and BALB/c mice (Fig. 1, Table 1). 3.2. Fertility features of NOD/SCID × NOD/SCID mice Compared with BALB/c mice, the NOD/SCID displayed roughly normal fertility features. No significant difference was observed between BALB/c × BALB/c and NOD/SCID × NOD/SCID mice in fertility features including the increase in maternal weight during pregnancy, the number of pups born per litter, median time interval from the setup of the breeding to the birth of the first litter and the duration of pregnancy (Table 2). However, the resorption rate of embryos was slightly higher in NOD/SCID × NOD/SCID mice (18.6%; 11 out of 59) than in BALB/c × BALB/c mice detected at day 13.5 of gestation (5.4%; 3 out of 56; n = 8 for each group, P = 0.041). The mean litter size provided by the supplier (Experimental Animal Center) for BALB/c was 5.8 ± 2.0 (number of litters = 92, number of pups = 536), which was not
44
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
Fig. 1. Flow cytometric analysis of CD45+ DX5+ cells in placenta from BALB/c (A1–A3) and non-obese diabetic (NOD)/severe combined immunodeficiency (SCID) mice (B1–B3). Left panel, gate set. R1, captured cells. Middle panel, double negative control. Right panel, cell percentage derived from cells gated in R1. The percentages of CD45+ DX5+ /CD45+ cells were significantly higher in BALB/c mice (71.67% in A3) than in NOD/SCID mice (6.00% in B3). FSC, forward scatter count.
significantly different from that in our BALB/c × BALB/c group (P = 0.688) or the PBS-treated NOD/SCID group (P = 0.784), but was significantly lower than that in the polyIC-treated group (P = 0.003) and higher than that in ASGM1-treated NOD/SCID mice (P < 0.001). Table 2 Fertility features among BALB/c and NOD/SCID mice treated with or without NK activator and inhibitor Pregnant mice
BALB/c
NOD/SCID
NOD/SCID
NOD/SCID
Treatment n Non-pregnant weight (g) Maximum weight (g) Weight increase (g) Percentage increasec Pup/litter Time to first litter (days) Duration of pregnancy (days)
PBS 8 22.2 ± 1.2 38.3 ± 3.0 16.1 ± 2.9 72.8 ± 5.1 6.1 ± 2.0 35.5 ± 6.8 19.5 ± 0.9
PBS 8 22.6 ± 1.3 39.4 ± 2.1 16.8 ± 3.1 75.1 ± 6.0 5.6 ± 1.7 35.4 ± 8.7 20.6 ± 1.3
polyIC 8 21.9 ± 1.2 44.0 ± 4.8a 22.1 ± 5.4a 101.8 ± 9.6a 8.1 ± 2.4a 37.1 ± 5.8 20.1 ± 1.1
ASGM1 8 22.5 ± 1.2 32.7 ± 4.7b 10.2 ± 5.0b 45.7 ± 8.4b 2.6 ± 1.9b 40.3 ± 8.4 20.4 ± 0.7
PBS, phosphate buffer solution; ASGMI, anti-asialo GM1; polyIC, polyinosinic-polycytidylic acid. a Student’s t-test: P < 0.05 vs. PBS-treated BALB/c and NOD/SCID mice. b Student’s t-test: P < 0.01 vs. PBS-treated BALB/c and NOD/SCID mice. c (Maximum weight-non-pregnant weight) non-pregnant weight × 100.
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
45
Table 3 Resorption rate of embryos and placental lymphocyte phenotyping Mice
BALB/c
NOD/SCID
NOD/SCID
NOD/SCID
P-value
Treatment Resorption rate (eight pairs/group) n CD45+ DX5+ CD94+ DX5+ CD94+ /DX5+
PBS 5.4% (3/56)
PBS 18.6% (11/59)
polyIC 12.3% (7/57)
ASGM1 39.2% (20/51)
– 0.041a , 0.231b , 0.032c
3 95.2 ± 1.5 61.8 ± 9.3 20.5 ± 5.2 45.1 ± 11.0
3 94.1 ± 0.6 15.6 ± 5.3 5.8 ± 2.3 17.5 ± 2.3
3 93.2 ± 1.7 47.4 ± 9.9 24.3 ± 3.6 93.5 ± 4.2
3 98.1 ± 1.9 0.7 ± 0.5 0.2 ± 0.2 0.0 ± 0.0
– 0.518a , 0.630b , 0.121c 0.013a , 0.047b , 0.050c 0.060a , 0.012b , 0.070c 0.071a , 0.000b , 0.002c
a b c
BALB/c (PBS-treated) vs. NOD/SCID (PBS-treated). NOD/SCID (PBS-treated) vs. NOD/SCID (polyIC-treated). NOD/SCID (PBS-treated) vs. NOD/SCID (ASGM1-treated).
3.3. Remnant NK cell in the NOD/SCID Although there is an NK cell deficit, a remnant NK cell group can be observed in placental lymphocytes isolated from the syngeneic mating of NOD/SCID × NOD/SCID mice (15.6 ± 5.3% on day 13.5 of gestation; Table 3). In addition, a certain level of NK cell activity can be detected by the effector-target lysis test in splenocytes from the NOD/SCID (Fig. 2). 3.4. Effects of polyIC administration on pregnancy outcomes Pregnancy outcomes of NOD/SCID mice could be improved by previous polyIC administration. The maternal weight increase during pregnancy and median number of pups born
Fig. 2. Natural killer (NK) cell activity of splenocytes against YAC-1 target cells. The activity was increased upon the stimulation with polyinosinic-polycytidylic acid (polyIC), and decreased after the administration of anti-asialo GM1 (ASGM1).
46
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
per litter were significantly increased in polyIC-treated NOD/SCID mice than in the control PBS group (101.8 ± 9.6% versus 75.1 ± 6.0%, P < 0.05; 8.1 ± 2.4 versus 5.6 ± 1.7, P < 0.05, respectively; Table 2). On the other hand, multiple injections of polyIC could decrease the resorption rate of embryos from 18.6 to 12.3%, but not reach the statistically supported level (P = 0.231, Table 3). 3.5. Effects of ASGM1 on pregnancy outcomes Both the maternal weight increase during pregnancy and the litter size were significantly smaller in ASGM1-treated NOD/SCID mice than in the control PBS group (45.7 ± 8.4% versus 75.1 ± 6.0%, P < 0.01; 2.6 ± 1.9 versus 5.6 ± 1.7, P < 0.01, respectively; Table 2). In addition, inhibition of NK cells by ASGM1 could boost the resorption rate on day 13.5 of gestation from 18.6% in the control PBS group to as high as 39.2% (P = 0.032) in NOD/SCID × NOD/SCID mice (Table 3). 3.6. Effects of polyIC administration on NK cell activity Upon stimulation with polyIC, the level of NK cell activity increased dramatically in splenocytes derived from NOD/SCID mice (see Fig. 2). In addition, the DX5+ , CD94+ and DX5+ CD94+ /DX5+ cell percentages in lymphocytes harvested from placentas were also significantly increased (47.4 ± 9.9% versus 15.6 ± 5.3%, P = 0.047; 24.3 ± 3.6% versus 5.8 ± 2.3%, P = 0.012; 93.5 ± 4.2% versus 17.5 ± 2.3%, P < 0.001, respectively; Figs. 3 and 4; Table 3). This was concomitant with the improved pregnancy outcomes (Tables 2 and 3). 3.7. Effects of ASGM1 on NK cell activity Natural killer cell activity derived from splenocytes was further inhibited after the multiple injections of ASGM1 anti-serum in NOD/SCID mice (Fig. 2). In addition, data from placenta showed that both DX5+ cell and DX5+ CD94+ cell percentages were dramatically decreased following the administration of ASGM1 (5.8 ± 2.3% versus 15.6 ± 5.3%, P = 0.050; 0.0 ± 0.0% versus 17.5 ± 2.3%, P = 0.002, respectively; Figs. 3 and 4; Table 3), concomitant with the increase in embryo resorption.
4. Discussion As the first step of this work, T cell, B cell and NK cell deficits were confirmed in NOD/SCID mice by flow cytometric analysis and YAC-1 cellular cytotoxicity assays. Previous histological examination by staining of CD45 confirmed a remarkable concentration of leukocytes at the implantation site on day 13.5 of pregnancy in normal mice (Parr et al., 1990). More than 95% of lymphocytes were detected to be CD45+ in NOD/SCID mice in another report using FCM (Shultz et al., 1995). A similar result was observed in our current study, supporting that CD45 can serve as an extensively expressed marker to distinguish lymphocytes from other cell groups, when gate thresholds are set adequately in FCM.
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
47
Fig. 3. Flow cytometric analysis of placenta NK cells. (A1–A3) BALB/c (phosphate buffer solution [PBS]-treated); (B1–B3) NOD/SCID (PBS-treated); (C1–C3) NOD/SCID (polyIC-treated); (D1–D3) NOD/SCID (ASGM1treated). Left panel, gate set. R1, captured cells. Middle panel, side scatter count (SSC)-FL1 (fluorescein isothiocyanate [FITC]). R2, phycoerythrin (PE)-stained cells. Right panel, DX5+ cells in the CD45+ cell group.
48
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
Fig. 4. Flow cytometric analysis of placenta DX5+ CD94+ cells. (A1–A3) BALB/c (PBS-treated); (B1–B3) NOD/SCID (PBS-treated); (C1–C3) NOD/SCID (polyIC-treated); (D1–D3) NOD/SCID (ASGM1-treated). Left panel, gate set. R1, captured cells. Middle panel, SSC-FL1 (FITC). R2, FITC-stained cells. Right panel, CD94+ cells in DX5+ cell group.
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
49
Although NOD/SCID mice exhibited NK cell deficit, a remnant NK cell activity was confirmed by effector-target lysis test in splenocytes both in this study and in a previous report (Christianson et al., 1997). In addition, a small percentage of CD45+ DX5+ NK cells could be detected with FCM in placental cells derived from NOD/SCID mice in the present study. Fertility features were observed in BALB/c ×BALB/c mice and NOD/SCID × NOD/SCID mice during a 4-month interval. In general, NOD/SCID mice displayed roughly normal pregnancy outcomes similar to BALB/c mice, strengthened by the animal supplier’s data on mean litter size for BALB/c, except for a slightly higher rate of embryo resorption on day 13.5 of gestation in NOD/SCID mice. However, these features could be dramatically changed by the administration of either polyIC or ASGM1. In this study, multiple injections of polyIC improved the pregnancy outcomes in NOD/SCID mice. In contrast, more embryos were resorbed after the administration of ASGM1. At the same time, as confirmed by FCM and cytotoxicity assay, the administration of polyIC had increased both the number of NK cells in placenta and the killing activity in splenocytes, while ASGM1 injection decreased both of them. These results indicated that adequate NK cell number and activity at the feto-maternal interface may be beneficial to the development of embryos in NOD/SCID mice. Cytotoxicity assays carried out in uNK cells in future will be a valuable addition to these results; they were not performed due to the limited number of NOD/SCID mice available. As for the influence of ASGM1 on uNK cells, some investigators did not find uterine depletion of NK cells, although they all reported splenic depletion (Guimond et al., 1997; Head, 1996–1997; Parr et al., 1990), but other reports (e.g., Clark et al., 1998) suggested that uNK cells could be depleted, or more accurately, inhibited by ASGM1 administration, and subsequently influence the pregnancy outcomes. Inhibition of uNK cells following ASGM1 injection was observed in our current study, supporting the report by Clark et al. (1998). It is not clear why there are outcome differences in the above-mentioned reports. However, on one hand, the extremely low level of basal NK cell proportion in immunodeficiency mice, such as the TgE26, may make it difficult to find the reduction in cell number after ASGM1 administration. On the other hand, flow cytometric analysis may be more sensitive to the change of uNK cell number than histological examination, while the latter is more valuable for localizing these cells at the feto-maternal interface. In the present study, the DX5+ CD94+ /DX5+ cell percentage in placenta was significantly boosted in NOD/SCID mice by polyIC administration, but dramatically decreased after ASGM1 treatment. These changes appeared to be associated with the alteration of embryo resorption. It implies that polyIC administration may have the potential to improve pregnancy outcomes of NOD/SCID by restoring the activity of remnant NK cells at the fetomaternal interface, which may be beneficial to embryo development. In contrast, inhibition of NK cells by ASGM1 could boost embryo resorption in NOD/SCID mice. It also suggests that sufficient expression of the CD94 molecule may be beneficial to the maintenance of embryos in NOD/SCID mice. However, the detailed mechanisms remain to be clarified in future studies. Under physiological conditions, NK cells are the dominant cell populations up to mid gestation in the pregnant uterus. Their contribution to pregnancy has been a controversial issue. For a long time, NK cells were considered to be candidate abortificient cells, but
50
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
King and Loke (1991) (Loke and King, 1991) pointed out that uNK cells may be essential to the establishment of blood supply for the feto-placental unit. In contrast, inhibition of NK cells was correlated with the decrease in resorption rate in abortion-prone CBA/J × DBA/2 mice. In the CBA/J × DBA/2 mice, it was believed that excessive NK cells infiltrated at the feto-maternal interface, and the frequency of such an infiltration was proportional to the resorption rate of embryos. Therefore, NK cell inhibition with ASGM1 was beneficial to pregnancy outcomes in these mice (Clark et al., 1998). In contrast to the observation in CBA/J × DBA/2 mice (Clark et al., 1998), ASGM1 administration, aimed at inhibiting NK cells in our current study, appeared to be harmful to embryo tolerance in NOD/SCID mice, a model of multiple immunodeficiency including NK cell deficit. Guimond et al. (1998) have demonstrated that NK cell deficits in TgE26 mice had profound reproductive defects, which could be cured by appropriate NK cell reconstitution. Our results indicate that the remnant NK cell activity is not harmful to fetomaternal immune tolerance in NOD/SCID mice. In contrast, a reduction in the remnant NK activity in these mice may be harmful to pregnancy outcomes. PolyIC, a synthesized double-strand RNA, is known to be an activator of NK cells. It was reported recently that in BALB/c macrophages, polyIC, as a stimulus of toll-like receptor 3 (TLR3), could strikingly induce the expression of a NKG2D ligand, RAE-1, in an innate immune response to infection (Hamerman et al., 2004). The potential role of polyIC may play in pregnancy tolerance and the detailed mechanisms remain to be investigated. Shimada et al. (2003) reported that embryo resorption could be boosted by polyIC in CBA/J × DBA/2 mice via the activation of NK cells and other cell groups. However, in our current study, multiple injections of polyIC during pregnancy appeared to be helpful to pregnancy tolerance in NK cell-deficient NOD/SCID mice, consistent with our above-mentioned results observed in ASGM1-treated NOD/SCID. It suggests that although excessive NK cell activity may be a disadvantage to the maintenance of embryos, inhibition of NK cells completely was also harmful to their development. In contrast, it may be helpful to pregnancy to increase the native low level of NK cells in NOD/SCID mice. Further research was restricted by the scale of NOD/SCID strains in our laboratory. Lymphocyte phenotyping can be performed at several time points during pregnancy, enabling the dynamic observation of NK cells and other cells, such as T cell and B cell subsets. Cytokines belonging to Th1 type (such as TNF-␣ and IFN-␥) and Th2 type (such as IL-10) can also be detected to describe the detailed mechanisms of pregnancy tolerance in an immunodeficient NOD/SCID model (Chaouat et al., 2004). In mice with multiple immunodeficiency, it is difficult to get a sufficient number of lymphocytes for FCM at an earlier stage of pregnancy (e.g., days 8.5–10.5). In future studies, it may be helpful to pool placentas harvested from several mice in the same group for flow cytometric analysis to describe the status of local immunity at an earlier stage of gestation. Acknowledgements This study was supported by the Postdoctoral Science Foundation of China (2002032244), the Nature and Science Foundation of Guangdong Province (4300213) and the Foundation for Outstanding Newcomers to Jinan University (51204067) (to Y. Lin).
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
51
References Arase, H., Saito, T., Phillips, J.H., Lanier, L.L., 2001. The mouse NK cell-associated antigen recognized by DX5 monoclonal antibody is CD49b (␣2 integrin, very late antigen-2). J. Immunol. 167, 1141–1144. Bosma, G.C., Custer, R.P., Bosma, M.J., 1983. A severe combined immunodeficiency mutation in the mouse. Nature 301, 527–530. Chaouat, G., Ledee-Bataille, N., Dubanchet, S., Zourbas, S., Sandra, O., Martel, J., 2004. Reproductive immunology 2003: reassessing the Th1/Th2 paradigm? Immunol. Lett. 92, 207–214. Christianson, S.W., Greiner, D.L., Hesselton, R., Leif, J.H., Wagar, E.J., Schweitzer, I.B., Rajan, T.V., Gott, B., Roopenian, D.C., Shultz, L.D., 1997. Enhanced human CD4+ T cell engraftment in 2 -microglobulin-deficient NOD-scid mice. J. Immunol. 158, 3578–3586. Christianson, S.W., Shultz, L.D., Leiter, E.H., 1993. Adoptive transfer of diabetes into immunodeficient NODscid/scid mice: relative contributions of CD4+ and CD8+ T lymphocytes from diabetic versus prediabetic NOD. NON-Thy-1a donors. Diabetes 42, 44–55. Clark, D.A., Chaouat, G., Arck, P.C., Mittruecker, H.W., Levy, G.A., 1998. Cytokine-dependent abortion in CBA × DBA/2 mice is mediated by the procoagulant fgl2 prothrombinase. J. Immunol. 160, 545–549. Dorshkind, K., Pollack, S.B., Bosma, M.J., Phillips, R.A., 1985. Natural killer (NK) cells are present in mice with severe combined immunodeficiency (scid). J. Immunol. 134, 3798–3801. Guimond, M.J., Luross, J.A., Wang, B., Terhorst, C., Danial, S., Croy, B.A., 1997. Absence of natural killer cells during murine pregnancy is associated with reproductive compromise in TgE26 mice. Biol. Reprod. 56, 169–179. Guimond, M.J., Wang, B., Croy, B.A., 1998. Engraftment of bone marrow from severe combined immunodeficiency (SCID) mice reverses the reproductive deficit in natural killer cells deficient TgE26 mice. J. Exp. Med. 187, 217–223. Hamerman, J.A., Ogasawara, K., Lanier, L.L., 2004. Toll-like receptor signaling in macrophages induces ligands for the NKG2D receptor. J. Immunol. 172, 2001–2005. Head, J.R., 1996–1997. Uterine natural killer cells during pregnancy in rodents. Nat. Immunol. 15, 7–21. Kataoka, S., Satoh, J., Fujiya, H., Toyota, T., Suzuki, R., Itoh, K., Kumagai, K., 1983. Immunologic aspects of the nonobese diabetic (NOD) mouse: abnormalities of cellular immunity. Diabetes 32, 247–253. King, A., Loke, Y.W., 1991. Uterine large granular lymphocytes: a possible role in embryonic implantation? Am. J. Obstet. Gynecol. 164, 702. Lee, I.F., Qin, H., Trudeau, J., Dutz, J., Tan, R., 2004. Regulation of autoimmune diabetes by complete Freund’s adjuvant is mediated by NK cells. J. Immunol. 172, 937–942. Lin, Y., Zeng, Y., Zhao, J., Zeng, S., Huang, J., Feng, Z., Di, J., Zhan, M., 2004. Murine CD45+ CD86+ cells isolated from para-aortic lymph nodes in an abortion-prone model. J. Reprod. Immunol. 64, 133–143. Loke, Y.W., King, A., 1991. Recent developments in the human maternal–fetal immune interaction. Curr. Opin. Immunol. 3, 762–766. Parr, E.L., Young, L.H., Parr, M.B., Young, J.D., 1990. Granulated metrial gland cells of pregnant mouse uterus are natural killer-like cells that contain perforin and serine esterases. J. Immunol. 145, 2365–2372. Shimada, S., Iwabuchi, K., Watano, K., Shimizu, H., Yamada, H., Minakami, H., Onoe, K., 2003. Expression of allograft inflammatory factor-1 in mouse uterus and poly (I:C)-induced fetal resorption. Am. J. Reprod. Immunol. 50, 104–112. Shultz, L.D., Schweitzer, P.A., Christianson, S.W., Gott, B., Schweitzer, I.B., Tennent, B., McKenna, S., Mobraaten, L., Rajan, T.V., Greiner, D.L., Leiter, E.H., 1995. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J. Immunol. 154, 180–191. Tempfer, C.B., Moreno, R.M., Gregg, A.R., 2000. Genetic control of fertility and embryonic waste in the mouse: a role for angiotensinogen. Biol. Reprod. 62, 457–462. Vance, R.E., Jamieson, A.M., Cado, D., Raulet, D.H., 2002. Implications of CD94 deficiency and monoallelic NKG2A expression for natural killer cell development and repertoire formation. Proc. Natl. Sci. U.S.A. 99, 868–873.
Lymphocyte phenotyping and NK cell activity analysis in pregnant NOD/SCID mice Yi Lin ∗ , Yijing Chen, Yaoying Zeng, Tong Wang, Shan Zeng Institute of Tissue Transplantation and Immunology, Jinan University, Guangzhou City 510632, China Received 1 April 2005; received in revised form 9 May 2005; accepted 9 May 2005
Abstract The purpose of this study was to outline the fertility features of non-obese diabetic (NOD)/LtSzscid/scid (NOD/SCID [severe combined immunodeficiency] for short) mice, and to evaluate the effects of NK cell subsets on the pregnancy outcomes of the syngeneic NOD/SCID × NOD/SCID mating combination. Firstly, lymphocyte phenotyping was performed with flow cytometry to confirm the multiple immunodeficits in NOD/SCID mice. Fertility features were assessed in NOD/SCID × NOD/SCID mice and compared with non-immunodeficiency control BALB/c × BALB/c mice. Although the presence of NK cell deficit is apparent in NOD/SCID mice, a certain level of remnant NK activity could be observed in these mice. The remnant NK cell activity was stimulated with polyinosinic-polycytidylic acid (polyIC), or inhibited with anti-asialo GM1 (ASGM1) anti-serum, respectively. The effects of these factors on pregnancy outcomes were evaluated after administration. Roughly normal fertility could be observed in NOD/SCID × NOD/SCID mice. However, a slightly larger sized litter was observed in polyIC-treated NOD/SCID × NOD/SCID mice compared with control NOD/SCID mice. In contrast, embryo resorption was boosted after ASGM1 injection, and correlated subsequently with a smaller litter size. It indicates that the remnant NK cell activity in NOD/SCID mice may be beneficial to feto-maternal tolerance during pregnancy. © 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Animal model; Immunodeficiency diseases; Placenta; Reproductive immunology
Abbreviations: ASGM1, anti-asialo GM1; FSC, forward scatter count; mAb, monoclonal antibody; NOD, non-obese diabetic mouse; polyIC, polyinosinic-polycytidylic acid; SCID, severe combined immunodeficiency; SSC, side scatter count; uNKcell, uterine NK cell ∗ Corresponding author. Tel.: +86 20 85220298; fax: +86 20 85221337. E-mail address: [email protected] (Y. Lin). 0165-0378/$ – see front matter © 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.jri.2005.05.002
40
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
1. Introduction Bosma et al. (1983) originally reported that C.B-17 mice homozygous for the severe combined immunodeficiency (scid/scid) mutation lack functional T and B lymphocytes. Although they are defective in cellular and humoral immune function, these mice display normal NK cell function (Dorshkind et al., 1985). On the other hand, the non-obese diabetes (NOD)/Lt strain is characterized by a functional deficit in natural killer (NK) cells (Kataoka et al., 1983) and tends to develop a high incidence of autoimmune insulin-dependent diabetes mellitus (IDDM; Christianson et al., 1993). Shultz et al. (1995) have established a murine model, NOD/LtSz-scid/scid (NOD/SCID [severe combined immunodeficiency] for short), characterized by multiple defects in innate immunity as well as in the absence of T and B cell function, by backcrossing of the scid mutation onto the NOD/Lt strain background. Although NK cell deficit is apparent in NOD/SCID mice, remnant NK cell activity could be detected in these mice (Christianson et al., 1997). To our knowledge, little is known about the remnant function of these cells in feto-maternal tolerance during pregnancy. In addition, as a result of T cell deficiency, NOD/SCID mice do not develop IDDM, which is mediated by T cells (Shultz et al., 1995). As early as 1991, King and Loke (1991) (Loke and King, 1991) had spotted that uterine NK (uNK) cells “may have a role in the control of implantation and the transformation of the uterine vasculature by trophoblast on which the blood supply to the feto-placental unit depends,” at a time when NK cells were only seen as candidate abortificient cells affecting mice. However, in CBA/J × DBA/2 mice, a widely used model of spontaneous abortion, activated NK cells and macrophages, was believed to be associated with the increase in embryo resorption (Clark et al., 1998). In the present study, T cell, B cell and NK cell subsets were analyzed with flow cytometry (FCM) to investigate whether there are any deficits in NOD/SCID mice. The fertility features were evaluated in these mice, and compared with a non-immunodeficiency model of BALB/c × BALB/c. The remnant NK cell activity in NOD/SCID mice was investigated and stimulated with polyinosinic-polycytidylic acid (polyIC), an activator of NK cells. In another design, inhibition of NK cells was performed by an i.p. injection of anti-asialo GM1 (ASGM1) anti-serum at an earlier stage of pregnancy in these mice. To our knowledge, CD94 is expressed mainly by NK cells and is present as a heterodimer with NKG2 on the cell surface. It plays an important role in the adhesion and activation of NK cell lineage (Vance et al., 2002). Therefore, the DX5+ CD94+ /DX5+ cell percentage in placental NK cells was determined with FCM using DX5 (its monoclonal antibody [mAb] is specific for CD49b) as a pan-NK cell marker (Arase et al., 2001). The effects of NK cells on pregnancy outcomes were evaluated in NOD/SCID × NOD/SCID mice.
2. Materials and methods 2.1. Mice BALB/c and NOD/LtSz-scid/scid mice (NOD/SCID for short) were purchased from the Experimental Animal Center of Zhongshan University (Guangzhou City, China), and bred
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
41
Table 1 Flow cytometric analysis of CD3+ , CD19+ and DX5+ cells with CD45 as a leukocyte common marker (n = 6 for each group) Cell source
Cell subset
Cell percentage
BALB/c
NOD/SCID
Peripheral blood Peripheral blood Placenta
T cell B cell NK cell
CD45+ CD3+ /CD45+ CD45+ CD19+ /CD45+ CD45+ DX5+ /CD45+
72.3 ± 6.8 3.6 ± 0.3 72.5 ± 3.2
4.3 ± 0.6a 1.3 ± 0.2a 6.4 ± 0.5a
NK, natural killer; NOD, non-obese diabetic mice; SCID, severe combined immunodeficiency. a Independent sample’s t-test: P < 0.001 vs. BALB/c group.
under specific pathogen-free conditions in the laboratory for immunodeficiency animals (Jinan University, Guangzhou, China). Each mouse was considered sexually mature, based on an age of 10–12 weeks. A syngeneic NOD/SCID × NOD/SCID breeding scheme was established, using the BALB/c × BALB/c mating combination as a non-immunodeficiency control. Lymphocyte phenotyping was performed in virgin mice and pregnant mice on day 13.5 of gestation. Glucosuria was assessed weekly in NOD/SCID mice for 4 months using Test strips (Baiyuan Mountain Pharmaceutical, Guangzhou, China). H&E-stained sections of thymus were examined to assess whether thymic lymphoma had developed in these mice. As expected, no urinary glucose was found in these mice, confirming their resistance to diabetes (Shultz et al., 1995). In addition, no thymic lymphoma was found in these mice at ages ranging from 12 to 30 weeks (Christianson et al., 1993; Shultz et al., 1995). Based on methods published previously (Shultz et al., 1995), we confirmed the T cell and B cell deficits in the NOD/SCID with FCM, using CD45, CD3 and CD19 as leukocyte common antigen, T cell marker and B cell marker, respectively. CD3+ cells in peripheral blood mononuclear cells (PBMC) were measured at 5.9% of the level found in control BALB/c mice, while CD19+ cells were 36.1% of the level in BALB/c, supporting the combined deficits in T cell and B cell subsets in the NOD/SCID (Table 1, P < 0.001). NK cell deficit in these mice was confirmed using placental cells and methods described in Sections 2.4 and 2.6. 2.2. Fertility features of NOD/SCID × NOD/SCID mice In the hope that pregnancy would result, one female was co-caged with one male, and on day 0.5 a vaginal plug was observed. Animals were housed under barrier conditions and monitored for health status. In addition, they were subjected to daily 12-h alternating periods of light and dark cycles within a controlled environment of 63% humidity at a temperature of 72–74 F. Food and water were provided ad libitum. Female mice were weighed twice a week. Data on each breeding pair were collected over a 4month period beginning with the establishment of the breeding (Tempfer et al., 2000). In another design, pregnant mice were sacrificed on day 13.5 of gestation. Peripheral blood, spleens, placentas and thymus were harvested. Collection of peripheral blood and placentas and purification of lymphocytes were performed using methods described in Section 2.6. The resorption rate of the embryos was calculated as follows: resorption
42
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
rate (%) = [number of resorbed embryos/number of total (resorbed + viable) embryos] × 100. 2.3. In vivo stimulation of NK cells Remnant NK cell activity was boosted by polyIC administration. Pregnant NOD/SCID mice were injected peritoneally with polyIC (␥-irradiated, Sigma Chemical, St. Louis, MO, USA), at a dosage of 100 g/20 g body weight (Shimada et al., 2003), while pregnant BALB/c and NOD/SCID mice injected with phosphate buffer solution (PBS) were used as controls. The injection was given on days 4.5, 7.5 and 10.5 of gestation consecutively. The efficacy of such treatment was evaluated by FCM using NK cell markers and YAC-1 cellular cytotoxicity assays, as described as follows (Lee et al., 2004; Shultz et al., 1995). 2.4. Assessment of NK cell activity The NK cell activity was determined as previously described (Lee et al., 2004; Shultz et al., 1995). Briefly, splenocytes were harvested from polyIC- and ASGM1-treated mice on day 13.5 of gestation, respectively. The target YAC-1 cells (American Type Culture Collection, Rockville, MD, USA) were labeled with 51 Cr and adjusted to 5 × 104 cells/ml. Various E:T ratios were set up in triplicate in V-bottom plates. Supernatants were harvested for counting on a gamma counter 4 h after incubation of targets and effectors at 37 ◦ C. The percentage-specific 51 Cr release was calculated using the following formula, where X is the mean experimental release from triplicate wells, total release (T) was determined from wells receiving 51 Cr-labeled YAC-1 target cells and 2% SDS, and spontaneous release (S) was determined from wells receiving 51 Cr-labeled YAC-1 target cells in growth medium: % specific release = [(X − S)/(T − S)] × 100. 2.5. In vivo inhibition of NK cells Inhibition of NK cells was performed using ASGM1 anti-serum (Wako Pure Chemical Inc., Osaka, Japan). NOD/SCID mice were injected peritoneally with 50 l of ASGM1 or an equivalent amount of PBS on days 4.5, 7.5 and 10.5 of gestation consecutively. The efficacy of such treatment was evaluated by FCM using NK cell markers and using standard YAC-1 cellular cytotoxicity assays, as mentioned above. 2.6. Monoclonal antibodies and flow cytometry The following mAbs were purchased from eBioscience (San Diego, CA, USA): fluorescein isothiocyanate (FITC)-conjugated CD3 (clone: 145-2C11); FITC- and phycoerythrin (PE)-conjugated CD45 (30-F11); FITC- and PE-conjugated DX5 (CD49b, clone: DX5); FITC-conjugated CD19 (MB19-1) and PE-conjugated CD94 (18d3). As previously described (Lin et al., 2004; Shultz et al., 1995), mononuclear cells were isolated from peripheral blood in virgin mice and peripheral blood and placentas in pregnant mice on day 13.5 of gestation. Briefly, peripheral blood was harvested from vena orbitalis, heparinized and purified by centrifugation on ficoll-hypaque density medium and incubated
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
43
with red cell lysis buffer. The isolated PBMCs were used for the confirmation of T cell and B cell deficits as mentioned above in Section 2.1. Hysterolaparotomy was performed to collect embryo-depleted placentas and associated decidual tissue, including decidua basalis. The pooled placentas were carefully cut into small pieces less than 1 mm3 with ocular scissors, and the pieces were then collected in Hanks’ balanced salt solution (HBSS) and filtered through a 50-m pore size nylon mesh to obtain mononuclear cell suspension. Mononuclear cells were purified by centrifugation of cell suspension on ficoll-hypaque density medium. Similarly, any contaminating red blood cells that might have persisted in the single-cell suspension were eliminated by incubation with red cell lysis buffer. Cells were incubated in the indicated mAb conjugates for 30 min in a total volume of 50 l of PBS containing 3% bovine serum albumen (BSA). After being washed twice with PBS, these cells were resuspended in 1% paraformaldehyde. Immunostained cells were analyzed on a FACSCalibur flow cytometer using CellQuest software (Becton Dickinson, San Jose, CA, USA). Ten thousand cells were detected in each sample. For a negative control, adequately designed gate thresholds were set to capture lymphoid cells using appropriate FITC- and PE-conjugated isotype control antibodies (eBioscience). 2.7. Statistical analysis A Student’s t-test was used to determine statistical significance where necessary. The resorption rate of embryos was analyzed using the χ2 -test.
3. Results 3.1. Confirmation of T, B and NK cell deficits in NOD/SCID mice In the current study, T cell and B cell deficits were confirmed by comparing the CD45+ CD3+ /CD45+ and CD45+ CD19+ /CD45+ cell percentages in PBMCs in nonpregnant NOD/SCID mice and BALB/c mice, respectively. NK cell deficit was confirmed by comparing the CD45+ DX5+ /CD45+ cell percentage in placental lymphocytes in NOD/SCID mice and BALB/c mice (Fig. 1, Table 1). 3.2. Fertility features of NOD/SCID × NOD/SCID mice Compared with BALB/c mice, the NOD/SCID displayed roughly normal fertility features. No significant difference was observed between BALB/c × BALB/c and NOD/SCID × NOD/SCID mice in fertility features including the increase in maternal weight during pregnancy, the number of pups born per litter, median time interval from the setup of the breeding to the birth of the first litter and the duration of pregnancy (Table 2). However, the resorption rate of embryos was slightly higher in NOD/SCID × NOD/SCID mice (18.6%; 11 out of 59) than in BALB/c × BALB/c mice detected at day 13.5 of gestation (5.4%; 3 out of 56; n = 8 for each group, P = 0.041). The mean litter size provided by the supplier (Experimental Animal Center) for BALB/c was 5.8 ± 2.0 (number of litters = 92, number of pups = 536), which was not
44
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
Fig. 1. Flow cytometric analysis of CD45+ DX5+ cells in placenta from BALB/c (A1–A3) and non-obese diabetic (NOD)/severe combined immunodeficiency (SCID) mice (B1–B3). Left panel, gate set. R1, captured cells. Middle panel, double negative control. Right panel, cell percentage derived from cells gated in R1. The percentages of CD45+ DX5+ /CD45+ cells were significantly higher in BALB/c mice (71.67% in A3) than in NOD/SCID mice (6.00% in B3). FSC, forward scatter count.
significantly different from that in our BALB/c × BALB/c group (P = 0.688) or the PBS-treated NOD/SCID group (P = 0.784), but was significantly lower than that in the polyIC-treated group (P = 0.003) and higher than that in ASGM1-treated NOD/SCID mice (P < 0.001). Table 2 Fertility features among BALB/c and NOD/SCID mice treated with or without NK activator and inhibitor Pregnant mice
BALB/c
NOD/SCID
NOD/SCID
NOD/SCID
Treatment n Non-pregnant weight (g) Maximum weight (g) Weight increase (g) Percentage increasec Pup/litter Time to first litter (days) Duration of pregnancy (days)
PBS 8 22.2 ± 1.2 38.3 ± 3.0 16.1 ± 2.9 72.8 ± 5.1 6.1 ± 2.0 35.5 ± 6.8 19.5 ± 0.9
PBS 8 22.6 ± 1.3 39.4 ± 2.1 16.8 ± 3.1 75.1 ± 6.0 5.6 ± 1.7 35.4 ± 8.7 20.6 ± 1.3
polyIC 8 21.9 ± 1.2 44.0 ± 4.8a 22.1 ± 5.4a 101.8 ± 9.6a 8.1 ± 2.4a 37.1 ± 5.8 20.1 ± 1.1
ASGM1 8 22.5 ± 1.2 32.7 ± 4.7b 10.2 ± 5.0b 45.7 ± 8.4b 2.6 ± 1.9b 40.3 ± 8.4 20.4 ± 0.7
PBS, phosphate buffer solution; ASGMI, anti-asialo GM1; polyIC, polyinosinic-polycytidylic acid. a Student’s t-test: P < 0.05 vs. PBS-treated BALB/c and NOD/SCID mice. b Student’s t-test: P < 0.01 vs. PBS-treated BALB/c and NOD/SCID mice. c (Maximum weight-non-pregnant weight) non-pregnant weight × 100.
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
45
Table 3 Resorption rate of embryos and placental lymphocyte phenotyping Mice
BALB/c
NOD/SCID
NOD/SCID
NOD/SCID
P-value
Treatment Resorption rate (eight pairs/group) n CD45+ DX5+ CD94+ DX5+ CD94+ /DX5+
PBS 5.4% (3/56)
PBS 18.6% (11/59)
polyIC 12.3% (7/57)
ASGM1 39.2% (20/51)
– 0.041a , 0.231b , 0.032c
3 95.2 ± 1.5 61.8 ± 9.3 20.5 ± 5.2 45.1 ± 11.0
3 94.1 ± 0.6 15.6 ± 5.3 5.8 ± 2.3 17.5 ± 2.3
3 93.2 ± 1.7 47.4 ± 9.9 24.3 ± 3.6 93.5 ± 4.2
3 98.1 ± 1.9 0.7 ± 0.5 0.2 ± 0.2 0.0 ± 0.0
– 0.518a , 0.630b , 0.121c 0.013a , 0.047b , 0.050c 0.060a , 0.012b , 0.070c 0.071a , 0.000b , 0.002c
a b c
BALB/c (PBS-treated) vs. NOD/SCID (PBS-treated). NOD/SCID (PBS-treated) vs. NOD/SCID (polyIC-treated). NOD/SCID (PBS-treated) vs. NOD/SCID (ASGM1-treated).
3.3. Remnant NK cell in the NOD/SCID Although there is an NK cell deficit, a remnant NK cell group can be observed in placental lymphocytes isolated from the syngeneic mating of NOD/SCID × NOD/SCID mice (15.6 ± 5.3% on day 13.5 of gestation; Table 3). In addition, a certain level of NK cell activity can be detected by the effector-target lysis test in splenocytes from the NOD/SCID (Fig. 2). 3.4. Effects of polyIC administration on pregnancy outcomes Pregnancy outcomes of NOD/SCID mice could be improved by previous polyIC administration. The maternal weight increase during pregnancy and median number of pups born
Fig. 2. Natural killer (NK) cell activity of splenocytes against YAC-1 target cells. The activity was increased upon the stimulation with polyinosinic-polycytidylic acid (polyIC), and decreased after the administration of anti-asialo GM1 (ASGM1).
46
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
per litter were significantly increased in polyIC-treated NOD/SCID mice than in the control PBS group (101.8 ± 9.6% versus 75.1 ± 6.0%, P < 0.05; 8.1 ± 2.4 versus 5.6 ± 1.7, P < 0.05, respectively; Table 2). On the other hand, multiple injections of polyIC could decrease the resorption rate of embryos from 18.6 to 12.3%, but not reach the statistically supported level (P = 0.231, Table 3). 3.5. Effects of ASGM1 on pregnancy outcomes Both the maternal weight increase during pregnancy and the litter size were significantly smaller in ASGM1-treated NOD/SCID mice than in the control PBS group (45.7 ± 8.4% versus 75.1 ± 6.0%, P < 0.01; 2.6 ± 1.9 versus 5.6 ± 1.7, P < 0.01, respectively; Table 2). In addition, inhibition of NK cells by ASGM1 could boost the resorption rate on day 13.5 of gestation from 18.6% in the control PBS group to as high as 39.2% (P = 0.032) in NOD/SCID × NOD/SCID mice (Table 3). 3.6. Effects of polyIC administration on NK cell activity Upon stimulation with polyIC, the level of NK cell activity increased dramatically in splenocytes derived from NOD/SCID mice (see Fig. 2). In addition, the DX5+ , CD94+ and DX5+ CD94+ /DX5+ cell percentages in lymphocytes harvested from placentas were also significantly increased (47.4 ± 9.9% versus 15.6 ± 5.3%, P = 0.047; 24.3 ± 3.6% versus 5.8 ± 2.3%, P = 0.012; 93.5 ± 4.2% versus 17.5 ± 2.3%, P < 0.001, respectively; Figs. 3 and 4; Table 3). This was concomitant with the improved pregnancy outcomes (Tables 2 and 3). 3.7. Effects of ASGM1 on NK cell activity Natural killer cell activity derived from splenocytes was further inhibited after the multiple injections of ASGM1 anti-serum in NOD/SCID mice (Fig. 2). In addition, data from placenta showed that both DX5+ cell and DX5+ CD94+ cell percentages were dramatically decreased following the administration of ASGM1 (5.8 ± 2.3% versus 15.6 ± 5.3%, P = 0.050; 0.0 ± 0.0% versus 17.5 ± 2.3%, P = 0.002, respectively; Figs. 3 and 4; Table 3), concomitant with the increase in embryo resorption.
4. Discussion As the first step of this work, T cell, B cell and NK cell deficits were confirmed in NOD/SCID mice by flow cytometric analysis and YAC-1 cellular cytotoxicity assays. Previous histological examination by staining of CD45 confirmed a remarkable concentration of leukocytes at the implantation site on day 13.5 of pregnancy in normal mice (Parr et al., 1990). More than 95% of lymphocytes were detected to be CD45+ in NOD/SCID mice in another report using FCM (Shultz et al., 1995). A similar result was observed in our current study, supporting that CD45 can serve as an extensively expressed marker to distinguish lymphocytes from other cell groups, when gate thresholds are set adequately in FCM.
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
47
Fig. 3. Flow cytometric analysis of placenta NK cells. (A1–A3) BALB/c (phosphate buffer solution [PBS]-treated); (B1–B3) NOD/SCID (PBS-treated); (C1–C3) NOD/SCID (polyIC-treated); (D1–D3) NOD/SCID (ASGM1treated). Left panel, gate set. R1, captured cells. Middle panel, side scatter count (SSC)-FL1 (fluorescein isothiocyanate [FITC]). R2, phycoerythrin (PE)-stained cells. Right panel, DX5+ cells in the CD45+ cell group.
48
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
Fig. 4. Flow cytometric analysis of placenta DX5+ CD94+ cells. (A1–A3) BALB/c (PBS-treated); (B1–B3) NOD/SCID (PBS-treated); (C1–C3) NOD/SCID (polyIC-treated); (D1–D3) NOD/SCID (ASGM1-treated). Left panel, gate set. R1, captured cells. Middle panel, SSC-FL1 (FITC). R2, FITC-stained cells. Right panel, CD94+ cells in DX5+ cell group.
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
49
Although NOD/SCID mice exhibited NK cell deficit, a remnant NK cell activity was confirmed by effector-target lysis test in splenocytes both in this study and in a previous report (Christianson et al., 1997). In addition, a small percentage of CD45+ DX5+ NK cells could be detected with FCM in placental cells derived from NOD/SCID mice in the present study. Fertility features were observed in BALB/c ×BALB/c mice and NOD/SCID × NOD/SCID mice during a 4-month interval. In general, NOD/SCID mice displayed roughly normal pregnancy outcomes similar to BALB/c mice, strengthened by the animal supplier’s data on mean litter size for BALB/c, except for a slightly higher rate of embryo resorption on day 13.5 of gestation in NOD/SCID mice. However, these features could be dramatically changed by the administration of either polyIC or ASGM1. In this study, multiple injections of polyIC improved the pregnancy outcomes in NOD/SCID mice. In contrast, more embryos were resorbed after the administration of ASGM1. At the same time, as confirmed by FCM and cytotoxicity assay, the administration of polyIC had increased both the number of NK cells in placenta and the killing activity in splenocytes, while ASGM1 injection decreased both of them. These results indicated that adequate NK cell number and activity at the feto-maternal interface may be beneficial to the development of embryos in NOD/SCID mice. Cytotoxicity assays carried out in uNK cells in future will be a valuable addition to these results; they were not performed due to the limited number of NOD/SCID mice available. As for the influence of ASGM1 on uNK cells, some investigators did not find uterine depletion of NK cells, although they all reported splenic depletion (Guimond et al., 1997; Head, 1996–1997; Parr et al., 1990), but other reports (e.g., Clark et al., 1998) suggested that uNK cells could be depleted, or more accurately, inhibited by ASGM1 administration, and subsequently influence the pregnancy outcomes. Inhibition of uNK cells following ASGM1 injection was observed in our current study, supporting the report by Clark et al. (1998). It is not clear why there are outcome differences in the above-mentioned reports. However, on one hand, the extremely low level of basal NK cell proportion in immunodeficiency mice, such as the TgE26, may make it difficult to find the reduction in cell number after ASGM1 administration. On the other hand, flow cytometric analysis may be more sensitive to the change of uNK cell number than histological examination, while the latter is more valuable for localizing these cells at the feto-maternal interface. In the present study, the DX5+ CD94+ /DX5+ cell percentage in placenta was significantly boosted in NOD/SCID mice by polyIC administration, but dramatically decreased after ASGM1 treatment. These changes appeared to be associated with the alteration of embryo resorption. It implies that polyIC administration may have the potential to improve pregnancy outcomes of NOD/SCID by restoring the activity of remnant NK cells at the fetomaternal interface, which may be beneficial to embryo development. In contrast, inhibition of NK cells by ASGM1 could boost embryo resorption in NOD/SCID mice. It also suggests that sufficient expression of the CD94 molecule may be beneficial to the maintenance of embryos in NOD/SCID mice. However, the detailed mechanisms remain to be clarified in future studies. Under physiological conditions, NK cells are the dominant cell populations up to mid gestation in the pregnant uterus. Their contribution to pregnancy has been a controversial issue. For a long time, NK cells were considered to be candidate abortificient cells, but
50
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
King and Loke (1991) (Loke and King, 1991) pointed out that uNK cells may be essential to the establishment of blood supply for the feto-placental unit. In contrast, inhibition of NK cells was correlated with the decrease in resorption rate in abortion-prone CBA/J × DBA/2 mice. In the CBA/J × DBA/2 mice, it was believed that excessive NK cells infiltrated at the feto-maternal interface, and the frequency of such an infiltration was proportional to the resorption rate of embryos. Therefore, NK cell inhibition with ASGM1 was beneficial to pregnancy outcomes in these mice (Clark et al., 1998). In contrast to the observation in CBA/J × DBA/2 mice (Clark et al., 1998), ASGM1 administration, aimed at inhibiting NK cells in our current study, appeared to be harmful to embryo tolerance in NOD/SCID mice, a model of multiple immunodeficiency including NK cell deficit. Guimond et al. (1998) have demonstrated that NK cell deficits in TgE26 mice had profound reproductive defects, which could be cured by appropriate NK cell reconstitution. Our results indicate that the remnant NK cell activity is not harmful to fetomaternal immune tolerance in NOD/SCID mice. In contrast, a reduction in the remnant NK activity in these mice may be harmful to pregnancy outcomes. PolyIC, a synthesized double-strand RNA, is known to be an activator of NK cells. It was reported recently that in BALB/c macrophages, polyIC, as a stimulus of toll-like receptor 3 (TLR3), could strikingly induce the expression of a NKG2D ligand, RAE-1, in an innate immune response to infection (Hamerman et al., 2004). The potential role of polyIC may play in pregnancy tolerance and the detailed mechanisms remain to be investigated. Shimada et al. (2003) reported that embryo resorption could be boosted by polyIC in CBA/J × DBA/2 mice via the activation of NK cells and other cell groups. However, in our current study, multiple injections of polyIC during pregnancy appeared to be helpful to pregnancy tolerance in NK cell-deficient NOD/SCID mice, consistent with our above-mentioned results observed in ASGM1-treated NOD/SCID. It suggests that although excessive NK cell activity may be a disadvantage to the maintenance of embryos, inhibition of NK cells completely was also harmful to their development. In contrast, it may be helpful to pregnancy to increase the native low level of NK cells in NOD/SCID mice. Further research was restricted by the scale of NOD/SCID strains in our laboratory. Lymphocyte phenotyping can be performed at several time points during pregnancy, enabling the dynamic observation of NK cells and other cells, such as T cell and B cell subsets. Cytokines belonging to Th1 type (such as TNF-␣ and IFN-␥) and Th2 type (such as IL-10) can also be detected to describe the detailed mechanisms of pregnancy tolerance in an immunodeficient NOD/SCID model (Chaouat et al., 2004). In mice with multiple immunodeficiency, it is difficult to get a sufficient number of lymphocytes for FCM at an earlier stage of pregnancy (e.g., days 8.5–10.5). In future studies, it may be helpful to pool placentas harvested from several mice in the same group for flow cytometric analysis to describe the status of local immunity at an earlier stage of gestation. Acknowledgements This study was supported by the Postdoctoral Science Foundation of China (2002032244), the Nature and Science Foundation of Guangdong Province (4300213) and the Foundation for Outstanding Newcomers to Jinan University (51204067) (to Y. Lin).
Y. Lin et al. / Journal of Reproductive Immunology 68 (2005) 39–51
51
References Arase, H., Saito, T., Phillips, J.H., Lanier, L.L., 2001. The mouse NK cell-associated antigen recognized by DX5 monoclonal antibody is CD49b (␣2 integrin, very late antigen-2). J. Immunol. 167, 1141–1144. Bosma, G.C., Custer, R.P., Bosma, M.J., 1983. A severe combined immunodeficiency mutation in the mouse. Nature 301, 527–530. Chaouat, G., Ledee-Bataille, N., Dubanchet, S., Zourbas, S., Sandra, O., Martel, J., 2004. Reproductive immunology 2003: reassessing the Th1/Th2 paradigm? Immunol. Lett. 92, 207–214. Christianson, S.W., Greiner, D.L., Hesselton, R., Leif, J.H., Wagar, E.J., Schweitzer, I.B., Rajan, T.V., Gott, B., Roopenian, D.C., Shultz, L.D., 1997. Enhanced human CD4+ T cell engraftment in 2 -microglobulin-deficient NOD-scid mice. J. Immunol. 158, 3578–3586. Christianson, S.W., Shultz, L.D., Leiter, E.H., 1993. Adoptive transfer of diabetes into immunodeficient NODscid/scid mice: relative contributions of CD4+ and CD8+ T lymphocytes from diabetic versus prediabetic NOD. NON-Thy-1a donors. Diabetes 42, 44–55. Clark, D.A., Chaouat, G., Arck, P.C., Mittruecker, H.W., Levy, G.A., 1998. Cytokine-dependent abortion in CBA × DBA/2 mice is mediated by the procoagulant fgl2 prothrombinase. J. Immunol. 160, 545–549. Dorshkind, K., Pollack, S.B., Bosma, M.J., Phillips, R.A., 1985. Natural killer (NK) cells are present in mice with severe combined immunodeficiency (scid). J. Immunol. 134, 3798–3801. Guimond, M.J., Luross, J.A., Wang, B., Terhorst, C., Danial, S., Croy, B.A., 1997. Absence of natural killer cells during murine pregnancy is associated with reproductive compromise in TgE26 mice. Biol. Reprod. 56, 169–179. Guimond, M.J., Wang, B., Croy, B.A., 1998. Engraftment of bone marrow from severe combined immunodeficiency (SCID) mice reverses the reproductive deficit in natural killer cells deficient TgE26 mice. J. Exp. Med. 187, 217–223. Hamerman, J.A., Ogasawara, K., Lanier, L.L., 2004. Toll-like receptor signaling in macrophages induces ligands for the NKG2D receptor. J. Immunol. 172, 2001–2005. Head, J.R., 1996–1997. Uterine natural killer cells during pregnancy in rodents. Nat. Immunol. 15, 7–21. Kataoka, S., Satoh, J., Fujiya, H., Toyota, T., Suzuki, R., Itoh, K., Kumagai, K., 1983. Immunologic aspects of the nonobese diabetic (NOD) mouse: abnormalities of cellular immunity. Diabetes 32, 247–253. King, A., Loke, Y.W., 1991. Uterine large granular lymphocytes: a possible role in embryonic implantation? Am. J. Obstet. Gynecol. 164, 702. Lee, I.F., Qin, H., Trudeau, J., Dutz, J., Tan, R., 2004. Regulation of autoimmune diabetes by complete Freund’s adjuvant is mediated by NK cells. J. Immunol. 172, 937–942. Lin, Y., Zeng, Y., Zhao, J., Zeng, S., Huang, J., Feng, Z., Di, J., Zhan, M., 2004. Murine CD45+ CD86+ cells isolated from para-aortic lymph nodes in an abortion-prone model. J. Reprod. Immunol. 64, 133–143. Loke, Y.W., King, A., 1991. Recent developments in the human maternal–fetal immune interaction. Curr. Opin. Immunol. 3, 762–766. Parr, E.L., Young, L.H., Parr, M.B., Young, J.D., 1990. Granulated metrial gland cells of pregnant mouse uterus are natural killer-like cells that contain perforin and serine esterases. J. Immunol. 145, 2365–2372. Shimada, S., Iwabuchi, K., Watano, K., Shimizu, H., Yamada, H., Minakami, H., Onoe, K., 2003. Expression of allograft inflammatory factor-1 in mouse uterus and poly (I:C)-induced fetal resorption. Am. J. Reprod. Immunol. 50, 104–112. Shultz, L.D., Schweitzer, P.A., Christianson, S.W., Gott, B., Schweitzer, I.B., Tennent, B., McKenna, S., Mobraaten, L., Rajan, T.V., Greiner, D.L., Leiter, E.H., 1995. Multiple defects in innate and adaptive immunologic function in NOD/LtSz-scid mice. J. Immunol. 154, 180–191. Tempfer, C.B., Moreno, R.M., Gregg, A.R., 2000. Genetic control of fertility and embryonic waste in the mouse: a role for angiotensinogen. Biol. Reprod. 62, 457–462. Vance, R.E., Jamieson, A.M., Cado, D., Raulet, D.H., 2002. Implications of CD94 deficiency and monoallelic NKG2A expression for natural killer cell development and repertoire formation. Proc. Natl. Sci. U.S.A. 99, 868–873.