Prednisolone in early pregnancy inhibits regulatory T cell generation and alters fetal and placental development in mice
Kieffer, Tom E C; Chin, Peck Y; Green, Ella S; Moldenhauer, Lachlan M; Prins, Jelmer R; Robertson, Sarah A
Published in:
Molecular human reproduction DOI:
10.1093/molehr/gaaa019
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Final author's version (accepted by publisher, after peer review)
Publication date: 2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Kieffer, T. E. C., Chin, P. Y., Green, E. S., Moldenhauer, L. M., Prins, J. R., & Robertson, S. A. (2020). Prednisolone in early pregnancy inhibits regulatory T cell generation and alters fetal and placental development in mice. Molecular human reproduction, 26(5), 340-352.
https://doi.org/10.1093/molehr/gaaa019
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Prednisolone in early pregnancy inhibits regulatory T cell generation and
1
alters fetal and placental development in mice
2 3
Tom E.C. Kieffer1,2, Peck Y. Chin1, Ella S. Green1, Lachlan M. Moldenhauer1, Jelmer R.
4
Prins2, and Sarah A. Robertson1*
5 6
1
Robinson Research Institute & Adelaide Medical School, University of Adelaide, Adelaide, 7
SA 5005, Australia 8
2
Department of Obstetrics and Gynaecology, University Medical Center Groningen, 9
University of Groningen, Groningen, The Netherlands 10
11
*Corresponding author: Sarah A. Robertson, The Robinson Research Institute, Adelaide
12
Medical School, University of Adelaide, Adelaide, SA 5005 Australia. E-mail: 13
sarah.robertson@adelaide.edu.au
14 15
Running title: Prednisolone disrupts maternal immune adaptation 16
17 18
Abstract
19
Corticosteroids have been utilised in the assisted reproduction setting with the expectation of 20
suppressing aberrant immune activation and improving fertility in women. However the 21
effects of corticosteroids on fertility, and on pregnancy and offspring outcomes, are unclear. 22
In this study, mice were administered prednisolone (1 mg/kg) or PBS daily in the pre-23
implantation phase, and effects on the adaptive immune response, the implantation rate, fetal 24
development, and postnatal outcomes were investigated. Prednisolone disrupted the expected 25
expansion of CD4+ T cells in early pregnancy, inhibiting generation of both regulatory T cells
26
(Treg cells) and effector T cells and suppressing IFNG required for T cell functional 27
competence. Prednisolone caused an 8-20% increase in the embryo implantation rate and 28
increased the number of viable pups per litter. In late gestation, fetal and placental weights 29
were reduced in a litter size-dependent manner, and the canonical inverse relationship 30
between litter size and fetal weight was lost. The duration of pregnancy was extended by ~0.5 31
day and birth weight was reduced by ~5% after prednisolone treatment. Viability of 32
prednisolone-exposed offspring was comparable to controls, but body weight was altered in 33
adulthood, particularly in male offspring. Thus, while prednisolone given in the pre-34
implantation phase in mice increases maternal receptivity to implantation and resource 35
investment in fetal growth, there is a trade-off in long-term consequences for fetal 36
development, birth weight and offspring health. These effects are associated with, and likely 37
caused by, prednisolone suppression of the adaptive immune response at the outset of 38
pregnancy. 39
40
Keywords: prednisolone, corticosteroids, immune suppression, immune tolerance, regulatory
41
T cells, implantation, fetal growth, fetal programming, parturition 42
43
Introduction
44
Prednisolone and other synthetic corticosteroids are regularly utilised off-label as an adjuvant 45
to assisted reproductive technology (ART) treatment in women, to mitigate implantation 46
failure and/or recurrent pregnancy loss (Kemp, et al., 2016, Robertson, et al., 2016, Hviid and 47
Macklon, 2017). The underlying justification is to suppress a potentially detrimental immune 48
response, particularly elevated uterine natural killer (uNK) cells, in order to promote embryo 49
survival and facilitate implantation (Quenby, et al., 2005, Ledee, et al., 2018, Cooper, et al., 50
2019). However, the biological and clinical rationale for prednisolone use in the absence of an 51
autoimmune diagnosis is debatable, given evidence that specific and well-regulated activation 52
of innate and adaptive immune cells in the endometrium, rather than immune suppression or 53
physical exclusion of immune cells, is favourable for successful implantation and pregnancy 54
success (Erlebacher, 2013, Dekel, et al., 2014, Robertson, et al., 2018). Moreover, substantial 55
clinical evidence shows a lack of effectiveness of prednisolone in unselected patient groups, 56
and questions the safety of its widespread use in reproductive medicine (Boomsma, et al., 57
2012, Ledee, et al., 2018, Cooper, et al., 2019). 58
Prednisolone acts through the constitutively-expressed glucocorticoid receptor (Vandevyver, 59
et al., 2014) to modulate immune cell phenotype and function, suppressing activation and 60
function particularly in T lymphocytes and also in natural killer cells, dendritic cells and 61
macrophages (Franchimont, 2004, Cain and Cidlowski, 2017). When present in appropriate 62
numbers and activation states, immune cells facilitate implantation through promoting 63
immune tolerance (Robertson, et al., 2018), supporting the decidual response and epithelial 64
receptivity to embryo attachment (Blois, et al., 2011, Mor, et al., 2011), and assisting 65
placental development through remodelling the decidual and uterine vasculature (Croy, et al., 66
2003, Care, et al., 2018). Cells of the adaptive immune compartment, and notably regulatory 67
T lymphocytes (Treg cells), are central regulators of the endometrial immune response 68
(Robertson, et al., 2018). Treg cells are essential for mediating maternal tolerance and anti-69
inflammatory effects at implantation, and modulate the phenotypes of uNK cells, 70
macrophages and dendritic cells (Guerin, et al., 2009, Robertson, et al., 2018). They also 71
promote uterine and decidual vascular dilatation and reduce oxidative stress in the placenta 72
(Cornelius, et al., 2015, Care, et al., 2018). Conversely, excessive T effector cells can 73
constrain implantation and cause fetal loss or fetal growth impairment (Xin, et al., 2014, 74
Crespo, et al., 2017, Moldenhauer, et al., 2017). 75
Here, we examine the effects of prednisolone in the peri-implantation phase on the maternal T 76
cell response, and analyse the consequences for pregnancy outcome and postnatal 77
development. We find that prednisolone acts to impair CD4+ T cell activation and particularly
78
Treg cell generation in early pregnancy. This is associated with elevated litter size and 79
reduced fetal and placental weight, loss of the canonical trade-off between litter size and fetal 80
weight, delayed timing of birth, and modest but significant changes in offspring growth and 81
development. 82
Materials and Methods
83
Mice and Treatments 84
CBA x C57Bl/6 F1 (CBA F1) female mice were purchased from the University of Adelaide 85
Animal Facility and BALB/c male mice from the Animal Resource Centre (Perth, Australia). 86
All mice were housed under specific pathogen-free conditions on a 12L:12D cycle, and food 87
and water were provided ad libitum. Experiments were performed in accordance with the 88
Australian Code of Practice for the Care and Use of Animals for Scientific Purposes, with 89
approval of the University of Adelaide Animal Ethics Committee. 90
Vaginal smears were taken daily from females at 8-12 weeks of age to monitor the estrous 91
cycle and pro-estrus females were placed with a BALB/c male. The next day, mice were 92
checked between 0800 h and 1000 h for presence of vaginal copulatory plugs, and the 93
morning of positive sighting was considered day 0.5 post coitum (pc). Mated females were 94
housed separately from males, with 2-4 females per cage, and randomly allocated to either the 95
prednisolone or control group. 96
On days 0.5 pc, 1.5 pc and 2.5 pc, mated females were administered either prednisolone 97
(Sigma, St Louis, USA) (20 µg in 100 µl 0.04% chloroform methanol (1:1) and 99.96% 98
sterile phosphate buffered saline (PBS)) or carrier control (100 µl 0.04% chloroform methanol 99
(1:1) and 99.96% sterile PBS) via intraperitoneal injection (ip) between 0800 h and 1000 h. 100
This dose of prednisolone (~0.8-1 mg/kg) was selected to reflect the higher end of the dose 101
range commonly used in ART treatment (5 mg to 60 mg per day (Boomsma, et al., 2012)). 102
Groups of mice were killed at estrus, and on days 3.5 and 9.5 pc for flow cytometry analysis 103
of CD4+ cells. Additional groups of mice were killed on day 5.5 pc for assessment of
104
implantation site numbers, on day 18.5 pc for analysis of late gestation fetal and placental 105
development, or progressed to term to determine perinatal outcomes and track post-natal 106
development of offspring until autopsy at 20 weeks of age. 107
Flow Cytometry 108
On day 3.5 pc or 9.5 pc, estrus or pregnant CBA F1 females were sedated with 2% 2.2.2-109
tribromoethanol (Avertin, Sigma), after which peripheral blood was collected by cardiac 110
puncture. Then, mice were killed by cervical dislocation and uterus-draining para-aortic 111
lymph nodes (PALN) and spleens were harvested. Blood was diluted in Hanks-buffered salt 112
solution (HBSS) and mononuclear leukocytes were isolated by centrifugation over 113
Lympholyte (Cedarlane Laboratories, Hornby, Canada) according to the manufacturer’s 114
instructions. Leukocytes were isolated from PALN by mechanical dispersion between glass 115
microscope slides. Spleens were crushed through a 70 µm cell strainer after which red blood 116
cells were lysed using red blood cell lysis buffer (154 mM NH4Cl, 99.9 mM KHCO3, 0.13
117
mM Na2EDTA diluted in 800 mL distilled H2O, pH 7.2).
118
Aliquots of 106 cells PALN and spleen cells were incubated (4 hours, 37°C, 5% CO2) in
119
phorbol 12-myristate 13-acetate (PMA; Sigma; 20 ng/mL), ionomycin (Invitrogen; 1 nM) and 120
BD Golgiplug (Brefeldin A; BD Biosciences, Franklin Lakes; 1 µl/mL as per the 121
manufacturer’s instructions) in RPMI for in-vitro stimulation. For live-dead staining, aliquots 122
of cells from tissues and blood were incubated (10 min, room temperature) with 0.001% 123
Fixable Viability Stain 620 (BD Biosciences) in PBS and thereafter washed with PBS. To 124
reduce non-specific binding, cells were incubated (15 min, room temperature) with anti-Fc-125
γIIR antibody (FcBlock; BD Biosciences). Subsequently cells were incubated (25 min, 4o
C) 126
with APC-Cy7 anti-CD4 (GK1.5, BD Biosciences) in 0.1% bovine serum albumin (BSA) in 127
0.05% sodium azide in PBS, pH 7.4 (fluorescence-activated cell sorting (FACS) buffer) and 128
washed with 0.04% sodium azide in PBS. After surface staining, cells were fixed and 129
permeabilised using a FOXP3 Staining Buffer set (eBioscience, USA) according to the 130
manufacturer’s instructions. Cells were incubated (25 min, 4°C) with APC anti-FOXP3 (FJK-131
16s, eBioscience), BV510 anti-IFNG (XMG1.2, Biolegend, USA), BV421 anti-CD304 132
(neuropilin-1, NRP1) (3E12, Biolegend), and PE anti-CTLA4 (UC10-4F10-11, BD 133
Biosciences) in permeablisation buffer (eBioscience). Cells were washed once with 134
permeabilisation buffer (eBioscience) and once with FACS buffer. Cells were then 135
resuspended in PBS and analysed using a FACSCantoII flow cytometer (BD Biosciences). 136
Analysis was performed using FlowJo v10 (LLC, USA). Cells stained with a single antibody 137
and fluorescence-minus-one controls were used for setting compensation and gates, 138
respectively (Tung, et al., 2004). 139
Pregnancy Outcome on Days 5.5 and 18.5 140
On day 5.5 pc or day 18.5 pc mice were killed by cervical dislocation between 1000 h and 141
1200 h. On day 5.5 pc, the number of implantation sites was counted. On day 18.5 pc the 142
number of viable and dead fetuses and resorption sites was counted. Total implantation count 143
was the sum of viable, dead and resorbing implantation sites. Viable pregnancy was defined 144
as the presence of at least one viable implantation site, and pregnancy rate was defined as the 145
proportion of mice with a vaginal plug that had progressed to viable pregnancy at day 18.5 pc. 146
Each viable fetus was dissected from the amniotic sac and umbilical cord. Fetal and placental 147
weights were recorded and the fetal:placental weight ratio was calculated. 148
Postnatal and Adult Offspring Outcome 149
At term, the time of spontaneous delivery was noted to the nearest 12 h, number of pups were 150
counted, and birth weight was recorded. Pups were weighed at 12-24 h, 8 days and 21 days 151
after birth, then weaned and housed with littermates of the same sex. All progeny were 152
weighed at 4 weeks, then every 2 weeks until 20 weeks of age. 153
Body Morphometry Analysis 154
At 20 weeks, a subset of progeny (1, or 2 if available, randomly selected male and female 155
progeny per litter) were killed by cervical dislocation, weighed and autopsied for full body 156
composition analysis. The following tissues were excised and weighed individually: brain, 157
heart, lungs (left and right), kidneys (left and right), liver, adrenal glands (left and right), 158
thymus, spleen, testes (males, left and right), seminal vesicle (males), epididymis (males), 159
ovaries (females, left and right), uterus (females), quadriceps (left and right), triceps (left and 160
right), biceps (left and right), gastrocnemius muscle (left and right), retroperitoneal fat, peri-161
renal fat, epididymal fat (males, left and right) and parametrial fat (females). Weights of 162
bilateral tissues and organs were combined. Total muscle weight is the sum of quadriceps, 163
triceps, biceps, and gastrocnemius muscles. Total central fat weight is the sum of 164
retroperitoneal fat, peri-renal fat and epididymal fat (for males) or parametrial fat (for 165
females), and the muscle:central fat ratio was determined. Total central fat weight was 166
subtracted from total body weight to calculate total lean weight. 167
Statistics and Data Analysis 168
Normality of distribution was assessed by Kolmogorov-Smirnov test. Normally distributed 169
data were analysed using Student T-test. Non-normally distributed data were analysed using 170
Mann-Whitney U test. Pregnancy rate and litter size distribution was analysed by 2-test.
171
Progeny weights, fetal and placental weights, and body composition data are expressed as 172
estimated marginal means ± SEM and were analysed by Mixed Model Linear Repeated 173
Measures ANOVA and post-hoc Sidak t-test, with dam as subject and litter size as covariate. 174
Within litter size categories, fetal and placental weights were compared by Sidak t-test. 175
Statistical analysis was conducted using SPSS version 20.0 software (SPSS Inc, Chicago, IL, 176
USA) and GraphPad Prism 5 software (GraphPad software Inc., San Diego, CA, USA). 177
Differences between treatment groups were considered significant when P < 0.05. 178
Results
180
Prednisolone reduces CD4+ T cells in early and mid-gestation pregnancy 181
To examine the effects of prednisolone administered in the peri-implantation phase on CD4+
182
T cells and Treg cells, the PALN from mice given prednisolone or carrier control were 183
analysed by flow cytometry at estrus (peri-ovulation), on day 3.5 pc (peri-implantation) and 184
day 9.5 pc (mid-gestation) (Fig. 1A). Control mice showed the expected dynamic changes in 185
CD4+ T cells (Fig. 1B) with a progressive increase in total CD4+ cells from estrus to
186
implantation and mid-gestation (Fig. 1C). CD4+ T cells were predominantly FOXP3
-187
conventional T (Tconv) cells (comprising T effector cells and naïve T cells) (Fig. 1D), 188
consistent with active involvement of the adaptive immune response in early pregnancy as 189
reported previously (van Mourik, et al., 2009, Mjosberg, et al., 2010, Mor, et al., 2011, 190
Moffett and Colucci, 2014). After prednisolone treatment, the expected increase in CD4+ T
191
cells failed to occur and their number remained similar across the time points (Fig. 1C, 1D). 192
Treg cells comprised 5-6% of CD4+ T cells on day 3.5 pc in control mice but were
193
differentially suppressed amongst CD4+ T cells after prednisolone treatment (Fig. 1E),
194
resulting in 37% fewer Treg cells at implantation, and 53% fewer Treg cells by day 9.5 (Fig. 195
1F). Thymic Treg (tTreg) cells and peripheral Treg (pTreg) cells, distinguished on the basis 196
of NRP1 expression (Yadav, et al., 2012), both contribute to the Treg increase in early 197
pregnancy (Moldenhauer, et al., 2019). Both tTreg and pTreg cells in the PALN failed to 198
increase after prednisolone treatment (Fig. 1G, 1H). Lymphocytes from spleen and blood 199
were also analysed, but were unchanged in prednisolone-treated mice (Supplemental Fig. 200
S1A-F). 201
Prednisolone alters T cell phenotype in PALN 202
IFNG is essential for the immunosuppressive function of Treg cells (Sawitzki, et al., 2005) 203
and effector functions of Tconv cells (Schoenborn and Wilson, 2007). Prednisolone did not 204
alter Treg cell IFNG expression on day 3.5 pc, but IFNG was lower in PALN Treg cells on 205
day 9.5 pc (Fig. 2A-C). Tconv cells expressing IFNG were also fewer on day 9.5 pc after 206
prednisolone treatment compared to controls (Fig. 2E, F). This difference was not seen in the 207
spleen or blood (Supplemental Fig. S2 A-F). 208
Prednisolone increases litter size 209
Pregnancy outcomes were evaluated in three cohorts of mated mice, assessed on day 5.5 pc, 210
day 18.5 pc or day 1 after birth. Prednisolone did not affect the likelihood of mated mice 211
progressing to pregnancy with healthy implantation sites or viable offspring (Fig. 3A). 212
Average litter sizes were increased 8.3%, 20.4% and 12.0% on day 5.5 pc, day 18.5 pc and 213
day 1 after birth respectively (P = 0.038 for day 18.5 pc) (Fig. 3B). Viable fetuses and 214
neonates per litter were similarly increased, with a 19.3% increase in viable implantation sites 215
at day 18.5 pc, and a 10.3% increase in viable pups on day 1 postpartum (P = 0.034 for day 216
18.5) (Fig. 3C). The proportion of implanted embryos retained as viable fetuses on day 18.5 217
pc was unaffected by prednisolone (95.2% versus 94.4% in control and prednisolone groups 218
respectively). Similarly the proportion of implantation sites resulting in viable offspring 219
surviving to day 18.5 pc, to live birth, or to weaning at 21 days was unaffected by maternal 220
prednisolone treatment (Fig. 3D). 221
Prednisolone exerted a striking effect on litter size distribution, with fewer litters of 6 or fewer 222
after prednisolone treatment in all three cohorts. Compared to control mice where 15%, 30% 223
and 29% of litters were 6 or fewer implantations sites or pups when measured on day 5.5 pc, 224
day 18.5 pc and post-partum day 1 respectively, only 6%, 5% and 0% of litters of 225
prednisolone-treated dams had 6 or fewer implants or pups (all P < 0.05) (Fig. 4A-4C). 226
Prednisolone treatment alters fetal and placental weights, and removes the fetal weight to 227
litter size trade-off 228
At day 18.5 pc, mixed model analysis of the effects of prednisolone treatment and litter size 229
showed a strong trend towards smaller fetal weights in dams given prednisolone (P = 0.057) 230
(Fig. 5A), while placental weight was significantly lower (P < 0.001) (Fig. 5B) and the 231
fetal:placental weight ratio (an indicator of placental efficiency) did not differ between the 232
prednisolone and control groups (Fig. 5C). Prednisolone was a significant determinant of fetal 233
weight, placental weight and fetal:placental weight ratio (all P < 0.001). An interaction 234
between prednisolone and litter size was a significant determinant of both fetal weight (P = 235
0.032) and placental weight (P = 0.035). 236
When data was analysed according to litter size category, the expected inverse relationship 237
between litter size and fetal weight (Maslennikova, et al., 2019) was evident in the control 238
group, but not in dams given prednisolone (Fig. 5D). The usual inverse relationship between 239
litter size and placental weight (Maslennikova, et al., 2019) was also lost (Fig. 5E), while the 240
fetal:placental ratio and litter size were positively correlated in controls, but not after 241
prednisolone treatment (Fig. 5F). Although the correlation coefficient values were influenced 242
by the higher proportion of small litters in control mice, the effects of prednisolone were still 243
evident when data from litter sizes smaller than 6 were excluded from the analysis 244
(Supplemental Fig. S3). Importantly, loss of the litter size-fetal weight relationship was 245
associated with notably smaller fetuses and placentas in moderate-sized litters comprising 7-9 246
fetuses in prednisolone-treated dams, while mean fetal and placental weights were similar 247
regardless of prednisolone in litters of 10 fetuses (Fig. 5D-5F). 248
Prednisolone treatment delays parturition and alters offspring post-natal development in a 249
sex-specific manner 250
Compared to controls, the average time of birth was delayed ~0.5 days in mice administered 251
prednisolone (P < 0.01) (Fig. 6A). On day 1 after birth, pups born to dams given prednisolone 252
were 9.4% lighter than pups of control dams (P < 0.01) (Fig. 6B). Lower weight in 253
prednisolone-exposed pups was still evident on day 8 after birth (P < 0.05) (Fig. 6B). 254
There was no effect of prednisolone on the proportion of male and female progeny weaned 255
(data not shown). Female progeny of prednisolone-treated dams had a significantly higher 256
weight compared to controls at 3 weeks (Fig. 6C). Thereafter until week 20, no difference in 257
weight of female progeny was observed. From 3 weeks until 16 weeks of age, there was no 258
effect of dam prednisolone treatment on male progeny weight (Fig. 6D), but at 18 weeks 259
males were lighter, with a trend towards lower weight also evident at 16 and 20 weeks (Fig. 260
6D). 261
At week 20, offspring were killed and body morphometry analysis was performed to evaluate 262
effects of prednisolone on individual organs and tissues (Suppl. Table I and II). In male 263
progeny of dams given prednisolone, adrenal weight was reduced by 18% (Suppl. Table I). 264
No other organs or tissues of male progeny showed effects of dam treatment (Suppl. Table I). 265
In female progeny, no differences in body morphometry were found after prednisolone 266
treatment (Suppl. Table II). 267
Discussion
268
In this study, we report that prednisolone administered to mice in the peri-implantation phase 269
disrupts maternal immune adaptation for pregnancy, preventing the expected rise in CD4+ T
270
cells, especially the critical CD4+FOXP3+ Treg cell subset. Although treatment caused an
271
increased litter size, there was a trade-off for offspring health, with fetal growth restriction, 272
delayed time of birth, reduced birth weight and altered growth trajectory of offspring. The 273
effect of prednisolone on litter size was exerted at the level of embryo implantation, with a 274
consistent 8-20% increase in the number of implantation sites, resulting in 10-19% more 275
viable offspring per litter in three independent cohorts. Given the well-described regulatory 276
influence of T cells on maternal receptivity to implantation and subsequent placental 277
development, we infer that the effects of prednisolone on litter size, and fetal and postnatal 278
growth, most likely involve a maternal immune mechanism. 279
In mice, as in other mammals, activation and proliferation of T cells occurs in the uterus-280
draining lymph nodes, in response to antigens and immune-regulatory agents delivered in 281
male seminal fluid at mating (Robertson, et al., 2018). In healthy pregnancy the PALN T cell 282
response is skewed towards Treg cells, due to selective expansion of tTreg cells derived from 283
the thymus, as well as pTregs induced from naïve T cells in the periphery (Moldenhauer, et 284
al., 2019). Treg cells then recirculate via the blood and are recruited into the endometrium to 285
facilitate embryo implantation (Aluvihare, et al., 2004, Shima, et al., 2010, Guerin, et al., 286
2011). A similar expansion in Treg cells occurs in early pregnancy in women, but is 287
diminished in many women experiencing implantation failure or recurrent pregnancy loss 288
(Sasaki, et al., 2004, Jasper, et al., 2006, Winger and Reed, 2011). 289
CD4+ T cells continue to increase over the course of pregnancy in control mice, but after
290
prednisolone treatment, they remain at non-pregnant levels through to mid-gestation. Treg 291
cells were more affected than Tconv cells at both early and mid-gestation time points. 292
Generation and function of Treg cells are known to be suppressed by prednisolone in other 293
tissue settings (Franchimont, 2004, Cain and Cidlowski, 2017). The relative impact of 294
prednisolone on different T cell subsets depends on the tissue context, exerting greater effects 295
on cells that are undergoing active proliferation (Stock, et al., 2005, Azab, et al., 2008, Wust, 296
et al., 2008, Sbiera, et al., 2011). In some conditions, corticosteroids shift the T cell balance to 297
favour Treg cells, particularly in humans or mice with auto-immune disorders (Chen, et al., 298
2006, Suarez, et al., 2006, Azab, et al., 2008, Luther, et al., 2009). In other conditions Treg 299
cells are selectively reduced, for example in mouse models of asthma and multiple sclerosis 300
(Stock, et al., 2005, Wust, et al., 2008), and in naïve mice (Sbiera, et al., 2011). The selective 301
effect of prednisolone on T cells and especially Treg cells in the PALN, but not the spleen or 302
blood, presumably reflects the nature and site of the T cell response to seminal fluid antigens 303
ongoing at that time. 304
The differential loss of Treg cells and shift towards T effector cells induced by prednisolone 305
might have been expected to suppress fertility. More extensive Treg cell depletion without 306
contemporaneous T effector cell removal has been shown to cause implantation failure and 307
disrupt placental development (Shima, et al., 2010, Teles, et al., 2013, Prins, et al., 2015, 308
Care, et al., 2018). Conversely, excessive T effector cells without counter-balancing Treg 309
cells elicit late gestation fetal loss (Xin, et al., 2014, Crespo, et al., 2017, Moldenhauer, et al., 310
2017). In the current study, Treg cells were reduced by ~40% at day 3.5 and ~60% at day 9.5, 311
while Tconv cells and pro-inflammatory IFNG+ T effector cells were also diminished but not
312
to the same degree. This reduced T effector cell pool presumably explains why fetal loss did 313
not occur. 314
Prednisolone strongly impacted IFNG production in CD4+ T cells. Although commonly 315
viewed as a pro-inflammatory cytokine, IFNG is also critical for Treg cell 316
immunosuppression (Sawitzki, et al., 2005, Koenecke, et al., 2012). Prednisolone has been 317
shown to inhibit IFNG expression in Treg cells in vitro (Daniel, et al., 2016). IFNG secretion 318
by Treg cells in gestational tissues influences the behaviour of macrophages and DCs, and 319
contributes to endometrial vasculature remodeling and maintenance of the decidua (Murphy, 320
et al., 2009, Dallagi, et al., 2015, Trojan, et al., 2016). 321
There is strong evidence from animal studies that prolonged or excessive exposure to natural 322
or synthetic glucocorticoids causes fetal growth impairment and programming of 323
cardiovascular, metabolic and neuroendocrine disorders (Seckl and Meaney, 2004, Kemp, et 324
al., 2016). Previous studies show that glucocorticoids administered in mid- or late gestation 325
affect placental development and gene expression in the placenta in a sex-dependent manner 326
(Vaughan, et al., 2015, Lee, et al., 2017, Ozmen, et al., 2017). The current study is the first to 327
demonstrate an effect of pre-implantation glucocorticoids on later fetal developmental 328
outcomes. A range of different mechanisms may contribute (Michael and Papageorghiou, 329
2008, Kemp, et al., 2016). Suppression of immune activation prior to implantation is a likely 330
underlying mechanism, given the key roles for Treg cells and other endometrial leukocytes in 331
uterine vascular adaptation and placental development (Care, et al., 2018). Several studies 332
show that disruption to immune adaptation in the pre-implantation phase can result in later 333
changes to placental development and fetal growth (Robertson, et al., 2018). 334
The glucocorticoid receptor is ubiquitously expressed (Vandevyver, et al., 2014) so 335
prednisolone may also exert effects in non-leukocytic lineages tissues in the endometrium, 336
including epithelial cells (Whirledge, et al., 2013) and endothelial cells (Henderson, et al., 337
2003), to modulate receptivity to implantation (Whirledge, et al., 2015). The embryo and 338
maternal tissues are particularly vulnerable to corticosteroids in the preimplantation phase, 339
since after implantation the fetus and placenta are better protected by placental 11βHSD2 340
(Fowden and Forhead, 2015). Altered epigenetic programming in the embryo might also play 341
a role (Drake, et al., 2007, Dunn, et al., 2010, Bale, 2011). 342
Unexpectedly, we found the extent of fetal growth restriction induced by prednisolone was 343
strongly related to litter size, with greater inhibitory effects in moderate-sized as opposed to 344
larger litters, and removal of the expected inverse relationship between litter size and fetal 345
weight. In life history theory, a canonical trade-off between litter number and offspring size is 346
a central tenet (Charnov and Morgan Ernest, 2006). This is achieved through maternally-347
imposed constraint on resource investment, in order to balance number and fitness of 348
offspring with maternal capacity for immediate and future reproductive success (Charnov and 349
Morgan Ernest, 2006). Selective female reproductive investment is well-described in 350
mammals, birds, reptiles and invertebrate species in response to internal and external drivers 351
that include male gamete quality, nutrition and environmental stressors (Chapman, 2006, 352
Eberhard, 2009). Our data indicate that prednisolone suppresses the female capacity to 353
selectively modulate litter size and fetal growth and the ability to appropriately manage the 354
relationship between them. This results in a moderate but significant cost to offspring fitness, 355
in terms of size at birth and altered postnatal growth. Additionally, extended gestation 356
represents a maternal cost. Like many programming insults, the effects of prednisolone 357
treatment during early pregnancy are sex-dependent. Male offspring born from dams exposed 358
to prednisolone before implantation showed lower weights in adulthood from 16 weeks after 359
birth, while female offspring transiently exhibited higher weights around weaning at 3 weeks. 360
The means by which prednisolone affects female control of litter size and fetal growth are 361
unclear, but our data point to an immune-mediated mechanism. Immune cells are particularly 362
sensitive to glucocorticoids, and immune-mediated modulation of homeostasis and function is 363
implicated as a key mechanism in many target tissues (Ramamoorthy and Cidlowski, 2016, 364
Cain and Cidlowski, 2017). The immune system is well-positioned to operate as an adaptive 365
response to environmental signals in female control of reproductive investment (Robertson, 366
2010, Robertson, et al., 2018). Previous studies in mice implicate immune signals including 367
maternal-fetal MHC disparity (Madeja, et al., 2011) and TNF expression (Maslennikova, et 368
al., 2019) in female control of implantation rate. Thus we speculate that prednisolone 369
suppresses a physiological mechanism through which the T cell population modulates the 370
number of implanting embryos, in order to match maternal resource availability. Further 371
studies are required to substantiate this, as clearly non-immune pathways influencing uterine 372
receptivity could also contribute (Michael and Papageorghiou, 2008, Whirledge, et al., 2015). 373
There is evidence of a trade-off between number and size of offspring in primates including 374
humans (Walker, et al., 2007), that may in part manifest through selective endometrial 375
receptivity at implantation (Teklenburg, et al., 2010). However, whether immune mediators 376
contribute to a selective response affecting resource investment at implantation in women, and 377
how these might be affected by prednisolone, remains unknown. 378
The increased gestation length in dams given prednisolone could reflect several factors. 379
Firstly, T cells responding to fetal antigens contribute to mechanisms of parturition (Gomez-380
Lopez, et al., 2013, Gomez-Lopez, et al., 2014). Failed expansion of the T cell pool after 381
prednisolone administration might result in fewer T effector cells available in late gestation to 382
accelerate parturition. Secondly, reduced fetal adrenal function might contribute, since fetal 383
glucocorticoid production is a key signal to initiate parturition (Li, et al., 2014). 384
Our observation of an effect on adrenal gland development is reminiscent of the effects of 385
antenatal glucocorticoid treatment on hypothalamic-pituitary-adrenal regulation (Gerardin, et 386
al., 2005, Waffarn and Davis, 2012). In rats, lower adrenal weight occurs in male offspring 387
exposed to stress-induced corticosterone (Gerardin, et al., 2005). It is known that later 388
gestation prednisolone can induce adrenal insufficiency (Broersen, et al., 2015), due to 389
perturbation of an immune-mediated developmental pathway (Tevosian, et al., 2015) or 390
epigenetic effects (Drake, et al., 2007, Bale, 2011). The consequences of an underdeveloped 391
adrenal gland can be significant; for example, adrenal insufficiency is linked with subfertility 392
in male rats (Gerardin, et al., 2005). Synthetic glucocorticoids administered in mid- or late 393
gestation can exert deleterious effects on development of organs including the kidney, brain 394
and heart (Drake, et al., 2007, Singh, et al., 2012). Although we did not find differences in 395
adult organ sizes or body morphometry other than the adrenal gland, this does not preclude 396
functional changes not examined in this study. 397
Increasing the chance of embryo implantation and pregnancy success is the goal of 398
administering prednisolone to women experiencing infertility. However our data raise the 399
question of whether prednisolone exacerbates immune dysfunction in some women, 400
especially when Treg cells are already reduced, as often occurs in women with implantation 401
failure or recurrent pregnancy loss (Sasaki, et al., 2004, Jasper, et al., 2006, Winger and Reed, 402
2011). This might explain why prednisolone is ineffective in unselected populations, as 403
shown in a Cochrane review (Boomsma, et al., 2012) and in more recent studies (Ledee, et al., 404
2018, Cooper, et al., 2019). There are insufficient data to draw conclusions on the impact of 405
corticosteroid use in ART on pregnancy and infant outcomes (Boomsma, et al., 2012), but 406
data from natural conceptions points to elevated miscarriage, preterm birth, stillbirth, fetal 407
growth restriction and fetal congenital anomalies (Reinisch, et al., 1978, Seckl and Meaney, 408
2004, Boomsma, et al., 2012, Hviid and Macklon, 2017). Cleft lip and palate and other 409
anomalies are associated with corticosteroid treatment in early pregnancy (Carmichael and 410
Shaw, 1999, Park-Wyllie, et al., 2000, Gur, et al., 2004). There is also evidence that 411
endogenous and exogenous glucocorticoids contribute to programming neuroendocrine, 412
metabolic and cardiovascular disorders in the human fetus (Seckl and Meaney, 2004). Our 413
experiments were not powered to detect fetal congenital anomalies so further studies are 414
required to investigate this aspect of the safety of prednisolone use in ART. 415
The dose and timing of prednisolone administered in this study is at the high end of the 5 mg 416
to 60 mg per day prednisolone treatment used in ART patients (Boomsma, et al., 2012), and is 417
likely to be allometrically higher again, given that prednisolone is delivered orally in humans. 418
Protocols for prednisolone use in ART generally involve administration from ovulation, 419
oocyte retrieval or the day of embryo transfer, for a period varying from three days up to three 420
months (Boomsma, et al., 2012), so three days of prednisolone administration prior to embryo 421
implantation in the current study recapitulates a short-term clinical regimen. Moreover there 422
are important differences between the mouse and human in terms of anatomy of the 423
reproductive tissues, and composition of the immune cells involved in embryo implantation 424
and placental development (Erlebacher, 2013, Robertson, et al., 2015), limiting the extent to 425
which the current study can be extrapolated to humans. Nevertheless, the results provide 426
insight on possible effects of prednisolone-induced immune suppression that need to be 427
evaluated in women. That even short-term use of prednisolone has long term consequences on 428
fetal outcomes in mice, raises the imperative to re-examine the rationale for, and safety of, 429
widespread clinical use of prednisolone in human reproductive medicine. 430
Authors' Roles:
431
S.A.R. designed and supervised the study. T.E.C.K., P.Y.C. and E.S.G. performed the 432
experiments. T.E.C.K., P.Y.C., E.S.G. and L.M.M. analysed the data. J.R.P. and L.M.M. 433
contributed to the project design and discussions on data interpretation. S.A.R and T.E.C.K. 434
wrote the manuscript. All authors revised drafts and reviewed the final manuscript. 435
436
Funding:
437
This study was supported by Project Grant APP1041335 (to S.A.R.) from the National Health 438
and Medical Research Council (Australia). 439 440 Conflict of Interest: 441 None declared. 442 443
References
444
Aluvihare VR, Kallikourdis M and Betz AG. Regulatory T cells mediate maternal tolerance to 445
the fetus. Nat Immunol 2004;5:266-271. 446
Azab NA, Bassyouni IH, Emad Y, Abd El-Wahab GA, Hamdy G and Mashahit MA. 447
CD4+CD25+ regulatory T cells (TREG) in systemic lupus erythematosus (SLE) patients: the 448
possible influence of treatment with corticosteroids. Clin Immunol 2008;127:151-157. 449
Bale TL. Sex differences in prenatal epigenetic programming of stress pathways. Stress 450
2011;14:348-356. 451
Blois SM, Klapp BF and Barrientos G. Decidualization and angiogenesis in early pregnancy: 452
unravelling the functions of DC and NK cells. J Reprod Immunol 2011;88:86-92. 453
Boomsma CM, Keay SD and Macklon NS. Peri-implantation glucocorticoid administration 454
for assisted reproductive technology cycles. Cochrane Database Syst Rev 2012;Cd005996. 455
Broersen LH, Pereira AM, Jorgensen JO and Dekkers OM. Adrenal Insufficiency in 456
Corticosteroids Use: Systematic Review and Meta-Analysis. J Clin Endocrinol Metab 457
2015;100:2171-2180. 458
Cain DW and Cidlowski JA. Immune regulation by glucocorticoids. Nat Rev Immunol 459
2017;17:233-247. 460
Care AS, Bourque SL, Morton JS, Hjartarson EP, Robertson SA and Davidge ST. Reduction 461
in regulatory T cells in early pregnancy causes uterine artery dysfunction in mice. 462
Hypertension 2018;72:177-187.
463
Carmichael SL and Shaw GM. Maternal corticosteroid use and risk of selected congenital 464
anomalies. Am J Med Genet 1999;86:242-244. 465
Chapman T. Evolutionary conflicts of interest between males and females. Current Biology 466
2006;16:R744-R754. 467
Charnov EL and Morgan Ernest SK. The offspring-size/clutch-size trade-off in mammals. The 468
American Naturalist 2006;167:578-582.
469
Chen X, Oppenheim JJ, Winkler-Pickett RT, Ortaldo JR and Howard OM. Glucocorticoid 470
amplifies IL-2-dependent expansion of functional FoxP3(+)CD4(+)CD25(+) T regulatory 471
cells in vivo and enhances their capacity to suppress EAE. Eur J Immunol 2006;36:2139-472
2149. 473
Cooper S, Laird SM, Mariee N, Li TC and Metwally M. The effect of prednisolone on 474
endometrial uterine NK cell concentrations and pregnancy outcome in women with 475
reproductive failure. A retrospective cohort study. J Reprod Immunol 2019;131:1-6. 476
Cornelius DC, Amaral LM, Harmon A, Wallace K, Thomas AJ, Campbell N, Scott J, Herse F, 477
Haase N, Moseley J, Wallukat G, Dechend R and LaMarca B. An increased population of 478
regulatory T cells improves the pathophysiology of placental ischemia in a rat model of 479
preeclampsia. Am J Physiol Regul Integr Comp Physiol 2015;309:R884-891. 480
Crespo AC, van der Zwan A, Ramalho-Santos J, Strominger JL and Tilburgs T. Cytotoxic 481
potential of decidual NK cells and CD8+ T cells awakened by infections. J Reprod Immunol 482
2017;119:85-90. 483
Croy BA, He H, Esadeg S, Wei Q, McCartney D, Zhang J, Borzychowski A, Ashkar AA, 484
Black GP, Evans SS, Chantakru S, van den Heuvel M, Paffaro VA, Jr. and Yamada AT. 485
Uterine natural killer cells: insights into their cellular and molecular biology from mouse 486
modelling. Reproduction 2003;126:149-160. 487
Dallagi A, Girouard J, Hamelin-Morrissette J, Dadzie R, Laurent L, Vaillancourt C, Lafond J, 488
Carrier C and Reyes-Moreno C. The activating effect of IFN-gamma on 489
monocytes/macrophages is regulated by the LIF-trophoblast-IL-10 axis via Stat1 inhibition 490
and Stat3 activation. Cell Mol Immunol 2015;12:326-341. 491
Daniel V, Trojan K and Opelz G. Immunosuppressive drugs affect induction of IFNy+ Treg 492
in vitro. Human Immunology 2016;77:146-152. 493
Dekel N, Gnainsky Y, Granot I, Racicot K and Mor G. The role of inflammation for a 494
successful implantation. Am J Reprod Immunol 2014;72:141-147. 495
Drake AJ, Tang JI and Nyirenda MJ. Mechanisms underlying the role of glucocorticoids in 496
the early life programming of adult disease. Clin Sci (Lond) 2007;113:219-232. 497
Dunn E, Kapoor A, Leen J and Matthews SG. Prenatal synthetic glucocorticoid exposure 498
alters hypothalamic-pituitary-adrenal regulation and pregnancy outcomes in mature female 499
guinea pigs. J Physiol 2010;588:887-899. 500
Eberhard WG. Postcopulatory sexual selection: Darwin's omission and its consequences. Proc 501
Natl Acad Sci U S A 2009;106 Suppl 1:10025-10032.
502
Erlebacher A. Immunology of the maternal-fetal interface. Annu Rev Immunol 2013;31:387-503
411. 504
Fowden AL and Forhead AJ. Glucocorticoids as regulatory signals during intrauterine 505
development. Exp Physiol 2015;100:1477-1487. 506
Franchimont D. Overview of the actions of glucocorticoids on the immune response: a good 507
model to characterize new pathways of immunosuppression for new treatment strategies. Ann 508
N Y Acad Sci 2004;1024:124-137.
509
Gerardin DC, Pereira OC, Kempinas WG, Florio JC, Moreira EG and Bernardi MM. Sexual 510
behavior, neuroendocrine, and neurochemical aspects in male rats exposed prenatally to 511
stress. Physiol Behav 2005;84:97-104. 512
Gomez-Lopez N, StLouis D, Lehr MA, Sanchez-Rodriguez EN and Arenas-Hernandez M. 513
Immune cells in term and preterm labor. Cell Mol Immunol 2014;11:571-581. 514
Gomez-Lopez N, Vega-Sanchez R, Castillo-Castrejon M, Romero R, Cubeiro-Arreola K and 515
Vadillo-Ortega F. Evidence for a role for the adaptive immune response in human term 516
parturition. Am J Reprod Immunol 2013;69:212-230. 517
Guerin LR, Moldenhauer LM, Prins JR, Bromfield JJ, Hayball JD and Robertson SA. Seminal 518
fluid regulates accumulation of FOXP3+ regulatory T cells in the preimplantation mouse 519
uterus through expanding the FOXP3+ cell pool and CCL19-mediated recruitment. Biol 520
Reprod 2011;85:397-408.
521
Guerin LR, Prins JR and Robertson SA. Regulatory T-cells and immune tolerance in 522
pregnancy: a new target for infertility treatment? Hum Reprod Update 2009;15:517-535. 523
Gur C, Diav-Citrin O, Shechtman S, Arnon J and Ornoy A. Pregnancy outcome after first 524
trimester exposure to corticosteroids: a prospective controlled study. Reprod Toxicol 525
2004;18:93-101. 526
Henderson TA, Saunders PT, Moffett-King A, Groome NP and Critchley HO. Steroid 527
receptor expression in uterine natural killer cells. J Clin Endocrinol Metab 2003;88:440-449. 528
Hviid MM and Macklon N. Immune modulation treatments-where is the evidence? Fertil 529
Steril 2017;107:1284-1293.
530
Jasper MJ, Tremellen KP and Robertson SA. Primary unexplained infertility is associated 531
with reduced expression of the T-regulatory cell transcription factor Foxp3 in endometrial 532
tissue. Mol Hum Reprod 2006;12:301-308. 533
Kemp MW, Newnham JP, Challis JG, Jobe AH and Stock SJ. The clinical use of 534
corticosteroids in pregnancy. Hum Reprod Update 2016;22:240-259. 535
Koenecke C, Lee CW, Thamm K, Fohse L, Schafferus M, Mittrucker HW, Floess S, Huehn J, 536
Ganser A, Forster R and Prinz I. IFN-gamma production by allogeneic Foxp3+ regulatory T 537
cells is essential for preventing experimental graft-versus-host disease. J Immunol 538
2012;189:2890-2896. 539
Ledee N, Prat-Ellenberg L, Petitbarat M, Chevrier L, Simon C, Irani EE, Vitoux D, Bensussan 540
A and Chaouat G. Impact of prednisone in patients with repeated embryo implantation 541
failures: Beneficial or deleterious? J Reprod Immunol 2018;127:11-15. 542
Lee JY, Yun HJ, Kim CY, Cho YW, Lee Y and Kim MH. Prenatal exposure to 543
dexamethasone in the mouse induces sex-specific differences in placental gene expression. 544
Dev Growth Differ 2017;59:515-525.
545
Li XQ, Zhu P, Myatt L and Sun K. Roles of glucocorticoids in human parturition: a 546
controversial fact? Placenta 2014;35:291-296. 547
Luther C, Adamopoulou E, Stoeckle C, Brucklacher-Waldert V, Rosenkranz D, Stoltze L, 548
Lauer S, Poeschel S, Melms A and Tolosa E. Prednisolone treatment induces tolerogenic 549
dendritic cells and a regulatory milieu in myasthenia gravis patients. J Immunol 550
2009;183:841-848. 551
Madeja Z, Yadi H, Apps R, Boulenouar S, Roper SJ, Gardner L, Moffett A, Colucci F and 552
Hemberger M. Paternal MHC expression on mouse trophoblast affects uterine vascularization 553
and fetal growth. Proc Natl Acad Sci U S A 2011;108:4012-4017. 554
Maslennikova SO, Gerlinskaya LA, Kontsevaya GV, Anisimova MV, Nedospasov SA, 555
Feofanova NA, Moshkin MP and Moshkin YM. TNFα is responsible for the canonical 556
offspring number-size trade-off. Sci Rep 2019;9:4568. 557
Michael AE and Papageorghiou AT. Potential significance of physiological and 558
pharmacological glucocorticoids in early pregnancy. Hum Reprod Update 2008;14:497-517. 559
Mjosberg J, Berg G, Jenmalm MC and Ernerudh J. FOXP3+ regulatory T cells and T helper 560
1, T helper 2, and T helper 17 cells in human early pregnancy decidua. Biol Reprod 561
2010;82:698-705. 562
Moffett A and Colucci F. Uterine NK cells: active regulators at the maternal-fetal interface. J 563
Clin Invest 2014;124:1872-1879.
564
Moldenhauer LM, Diener KR, Hayball JD and Robertson SA. An immunogenic phenotype in 565
paternal antigen-specific CD8+ T cells at embryo implantation elicits later fetal loss in mice. 566
Immunol Cell Biol 2017;95:705-715.
567
Moldenhauer LM, Schjenken JE, Hope CM, Green ES, Zhang B, Eldi P, Hayball JD, Barry 568
SC and Robertson SA. Thymus-derived regulatory T cells exhibit Foxp3 epigenetic 569
modification and phenotype attenuation after mating in mice. J Immunol 2019;203:647-657. 570
Mor G, Cardenas I, Abrahams V and Guller S. Inflammation and pregnancy: the role of the 571
immune system at the implantation site. Ann N Y Acad Sci 2011;1221:80-87. 572
Murphy SP, Tayade C, Ashkar AA, Hatta K, Zhang J and Croy BA. Interferon gamma in 573
successful pregnancies. Biol Reprod 2009;80:848-859. 574
Ozmen A, Unek G and Korgun ET. Effect of glucocorticoids on mechanisms of placental 575
angiogenesis. Placenta 2017;52:41-48. 576
Park-Wyllie L, Mazzotta P, Pastuszak A, Moretti ME, Beique L, Hunnisett L, Friesen MH, 577
Jacobson S, Kasapinovic S, Chang D, Diav-Citrin O, Chitayat D, Nulman I, Einarson TR and 578
Koren G. Birth defects after maternal exposure to corticosteroids: prospective cohort study 579
and meta-analysis of epidemiological studies. Teratology 2000;62:385-392. 580
Prins JR, Zhang B, Schjenken JE, Guerin LR, Barry SC and Robertson SA. Unstable Foxp3+ 581
regulatory T cells and altered dendritic cells are associated with lipopolysaccharide-induced 582
fetal loss in pregnant interleukin 10-deficient mice. Biol Reprod 2015;93:95. 583
Quenby S, Kalumbi C, Bates M, Farquharson R and Vince G. Prednisolone reduces 584
preconceptual endometrial natural killer cells in women with recurrent miscarriage. Fertil 585
Steril 2005;84:980-984.
586
Ramamoorthy S and Cidlowski JA. Corticosteroids: mechanisms of action in health and 587
disease. Rheum Dis Clin North Am 2016;42:15-31, vii. 588
Reinisch JM, Simon NG, Karow WG and Gandelman R. Prenatal exposure to prednisone in 589
humans and animals retards intrauterine growth. Science 1978;202:436-438. 590
Robertson SA. Immune regulation of conception and embryo implantation-all about quality 591
control? J Reprod Immunol 2010;85:51-57. 592
Robertson SA, Care AS and Moldenhauer LM. Regulatory T cells in embryo implantation and 593
the immune response to pregnancy. J Clin Invest 2018;128:4224-4235. 594
Robertson SA, Jin M, Yu D, Moldenhauer LM, Davies MJ, Hull ML and Norman RJ. 595
Corticosteroid therapy in assisted reproduction - immune suppression is a faulty premise. 596
Hum Reprod 2016;31:2164-2173.
597
Robertson SA, Petroff MG and Hunt JS. Immunology of Pregnancy. In Plant, T. M. and 598
Zeleznik, A. J. (eds), Knobil and Neill's Physiology of Reproduction. 2015; Elsevier B.V., 599
Amsterdam. 600
Sasaki Y, Sakai M, Miyazaki S, Higuma S, Shiozaki A and Saito S. Decidual and peripheral 601
blood CD4+CD25+ regulatory T cells in early pregnancy subjects and spontaneous abortion 602
cases. Mol Hum Reprod 2004;10:347-353. 603
Sawitzki B, Kingsley CI, Oliveira V, Karim M, Herber M and Wood KJ. IFN-gamma 604
production by alloantigen-reactive regulatory T cells is important for their regulatory function 605
in vivo. J Exp Med 2005;201:1925-1935. 606
Sbiera S, Dexneit T, Reichardt SD, Michel KD, van den Brandt J, Schmull S, Kraus L, Beyer 607
M, Mlynski R, Wortmann S, Allolio B, Reichardt HM and Fassnacht M. Influence of short-608
term glucocorticoid therapy on regulatory T cells in vivo. PLoS One 2011;6:e24345. 609
Schoenborn JR and Wilson CB. Regulation of interferon-gamma during innate and adaptive 610
immune responses. Adv Immunol 2007;96:41-101. 611
Seckl JR and Meaney MJ. Glucocorticoid programming. Ann N Y Acad Sci 2004;1032:63-84. 612
Shima T, Sasaki Y, Itoh M, Nakashima A, Ishii N, Sugamura K and Saito S. Regulatory T 613
cells are necessary for implantation and maintenance of early pregnancy but not late 614
pregnancy in allogeneic mice. J Reprod Immunol 2010;85:121-129. 615
Singh RR, Cuffe JS and Moritz KM. Short- and long-term effects of exposure to natural and 616
synthetic glucocorticoids during development. Clin Exp Pharmacol Physiol 2012;39:979-989. 617
Stock P, Akbari O, DeKruyff RH and Umetsu DT. Respiratory tolerance is inhibited by the 618
administration of corticosteroids. J Immunol 2005;175:7380-7387. 619
Suarez A, Lopez P, Gomez J and Gutierrez C. Enrichment of CD4+ CD25high T cell 620
population in patients with systemic lupus erythematosus treated with glucocorticoids. Ann 621
Rheum Dis 2006;65:1512-1517.
622
Teklenburg G, Salker M, Molokhia M, Lavery S, Trew G, Aojanepong T, Mardon HJ, 623
Lokugamage AU, Rai R, Landles C, Roelen BA, Quenby S, Kuijk EW, Kavelaars A, Heijnen 624
CJ, Regan L, Brosens JJ and Macklon NS. Natural selection of human embryos: decidualizing 625
endometrial stromal cells serve as sensors of embryo quality upon implantation. PLoS ONE 626
2010;5:e10258. 627
Teles A, Schumacher A, Kuhnle MC, Linzke N, Thuere C, Reichardt P, Tadokoro CE, 628
Hammerling GJ and Zenclussen AC. Control of uterine microenvironment by Foxp3(+) cells 629
facilitates embryo implantation. Front Immunol 2013;4:158. 630
Tevosian SG, Jimenez E, Hatch HM, Jiang T, Morse DA, Fox SC and Padua MB. Adrenal 631
Development in Mice Requires GATA4 and GATA6 Transcription Factors. Endocrinology 632
2015;156:2503-2517. 633
Trojan K, Unterrainer C, Aly M, Zhu L, Weimer R, Bulut N, Morath C, Opelz G and Daniel 634
V. IFNy+ and IFNy- Treg subsets with stable and unstable Foxp3 expression in kidney 635
transplant recipients with good long-term graft function. Transpl Immunol 2016. 636
Tung JW, Parks DR, Moore WA, Herzenberg LA and Herzenberg LA. New approaches to 637
fluorescence compensation and visualization of FACS data. Clin Immunol 2004;110:277-283. 638
van Mourik MS, Macklon NS and Heijnen CJ. Embryonic implantation: cytokines, adhesion 639
molecules, and immune cells in establishing an implantation environment. J Leukoc Biol 640
2009;85:4-19. 641
Vandevyver S, Dejager L and Libert C. Comprehensive overview of the structure and 642
regulation of the glucocorticoid receptor. Endocr Rev 2014;35:671-693. 643
Vaughan OR, Fisher HM, Dionelis KN, Jeffreys EC, Higgins JS, Musial B, Sferruzzi-Perri 644
AN and Fowden AL. Corticosterone alters materno-fetal glucose partitioning and insulin 645
signalling in pregnant mice. J Physiol 2015;593:1307-1321. 646
Waffarn F and Davis EP. Effects of antenatal corticosteroids on the hypothalamic-pituitary-647
adrenocortical axis of the fetus and newborn: experimental findings and clinical 648
considerations. Am J Obstet Gynecol 2012;207:446-454. 649
Walker RS, Gurven M, Burger O and Hamilton MJ. The trade-off between number and size 650
of offspring in humans and other primates. Proc Biol Sci 2007;275:827-833. 651
Whirledge S, Xu X and Cidlowski JA. Global gene expression analysis in human uterine 652
epithelial cells defines new targets of glucocorticoid and estradiol antagonism. Biol Reprod 653
2013;154:499-510. 654
Whirledge SD, Oakley RH, Myers PH, Lydon JP, DeMayo F and Cidlowski JA. Uterine 655
glucocorticoid receptors are critical for fertility in mice through control of embryo 656
implantation and decidualization. Proc Natl Acad Sci U S A 2015. 657
Winger EE and Reed JL. Low circulating CD4(+) CD25(+) Foxp3(+) T regulatory cell levels 658
predict miscarriage risk in newly pregnant women with a history of failure. Am J Reprod 659
Immunol 2011;66:320-328.
660
Wust S, van den Brandt J, Tischner D, Kleiman A, Tuckermann JP, Gold R, Luhder F and 661
Reichardt HM. Peripheral T cells are the therapeutic targets of glucocorticoids in 662
experimental autoimmune encephalomyelitis. J Immunol 2008;180:8434-8443. 663
Xin L, Ertelt JM, Rowe JH, Jiang TT, Kinder JM, Chaturvedi V, Elahi S and Way SS. Cutting 664
edge: committed Th1 CD4+ T cell differentiation blocks pregnancy-induced Foxp3 665
expression with antigen-specific fetal loss. J Immunol 2014;192:2970-2974. 666
Yadav M, Louvet C, Davini D, Gardner JM, Martinez-Llordella M, Bailey-Bucktrout S, 667
Anthony BA, Sverdrup FM, Head R, Kuster DJ, Ruminski P, Weiss D, Von Schack D and 668
Bluestone JA. Neuropilin-1 distinguishes natural and inducible regulatory T cells among 669
regulatory T cell subsets in vivo. J Exp Med 2012;209:1713-1722. 670
671
Figure 1. The effect of pre-implantation prednisolone on proportions and total cell
672
number of CD4+ T cells, Treg cells, thymic Treg cells, peripheral Treg cells, and Tconv
673
cells in the uterus-draining para-aortic lymph nodes (PALN) in early and mid-gestation.
674
PALN from CBA F1 females in estrus (est), on day 3.5 post coitum (pc), and on day 9.5 pc 675
after mating with BALB/c males, were analysed by flow cytometry. Representative dot plots 676
show gating of CD4+ T cells, Tconv cells (CD4+FOXP3-) and Treg cells (CD4+FOXP3+), as
677
well as thymic Treg cells (CD4+FOXP3+NRP1+) and peripheral Treg cells
678
(CD4+FOXP3+NRP1-), in prednisolone-treated and control mice on day 3.5 pc (A). Data are
679
the mean ± SEM proportion of CD4+ T cells (%live) (B), total CD4+ cell numbers (C),
680
proportion of Treg cells (%CD4+) (D), total Treg cell numbers (E), total thymic Treg cell
681
numbers (F), total peripheral Treg cell numbers (G), and total Tconv cell numbers (H) in the 682
PALN of estrus mice (closed squares) or mated mice treated with prednisolone (open squares) 683
or carrier control (closed circle) (n = 5-12 mice per group, per time point, per treatment). 684
Treatment effects were analysed by Mann-Whitney U-test (mean ± SEM) (*P < 0.05, **P < 685
0.01 and #P < 0.1).
686
687 688 689 690 691 692 693 694
Figure 2. The effect of pre-implantation prednisolone on IFNG+ T regulatory (Treg) 695
cells and Tconv cells in the uterus-draining para-aortic lymph nodes (PALN) in early
696
and mid-gestation. PALN of estrus mice (est), or mated mice on day 3.5 post coitum (pc),
697
and on day 9.5 pc, were analysed by flow cytometry. Representative dot plots show gating of 698
IFNG+ Treg cells (CD4+FOXP3+) and IFNG+ Tconv cells (CD4+FOXP3-), in
prednisolone-699
treated and control mice on day 3.5 pc (A). Data are the mean ± SEM proportion of IFNG+
700
Treg cells (%CD4+FOXP3+) (B) and total number of IFNG+ Treg cells (CD4+FOXP3+IFNG+)
701
(C) and proportion of IFNG+ Tconv cells (%CD4+FOXP3-) (D) and total number of IFNG+ 702
Tconv cells (CD4+FOXP3-IFNG+) (E) in the PALN of estrus mice (closed squares) or mated
703
mice treated with prednisolone (open squares) or carrier control (closed circle) (n = 5-12 mice 704
per group, per time point, per treatment). Treatment effects were analysed by Mann-Whitney 705
U-test (*P < 0.05, **P < 0.01). 706
707
708
Figure 3. The effects of peri-implantation prednisolone treatment on pregnancy rate,
709
total litter size, number of viable fetuses (or pups) per litter, and fetal (or pup) survival.
710
Pregnancy rate was the proportion of mated mice with at least one viable implantation site, 711
fetus or pup on day 5.5 post coitum (pc), day 18.5 pc, and day 1 post partum (A). Other data 712
are the mean ± SEM total litter size (B), number of viable fetuses / pups per litter (C), and 713
fetal / pup survival (D) counted on day 5.5 pc and 18.5 pc. Fetal / pup survival is viable 714
fetuses or pups as a percentage of implantation sites on day 5.5 pc in mice treated with 715
prednisolone (open squares) or carrier control (closed circles) (n = 20 mated mice per group, 716
per time point, per treatment). Treatment effects were analysed by 2-test for pregnancy rate
717
and Sidak t-test for litter size and fetal / pup viability (*P < 0.05). 718
719
Figure 4. The effect of peri-implantation prednisolone treatment on litter size
720
distribution. The proportion of total litters with 2-11 implantation sites or 2-10 total pups in
721
cohorts of dams treated with prednisolone (striped bars) or carrier control (open bars) 722
analysed on day 5.5 post coitum (pc) (A), day 18.5 pc (B), or day 1 post partum (C) (n = 20 723
mated mice per group, per time point, per treatment). Effects of treatment group on litter size 724
distribution were analysed by 2-test. Prednisolone treatment caused a significant shift to
725
larger litter sizes in all three cohorts (*P < 0.05). 726
727 728 729 730
Figure 5. The effects of peri-implantation prednisolone treatment on fetal and placental
731
weight analysed on day 18.5 post coitum (pc). Fetal weight (A), placental weight (B), and
732
fetal:placental weight ratio (C) in dams treated with prednisolone (striped bars) or carrier 733
control (open bars) are shown as estimated marginal mean ± SEM. Numbers of mated mice 734
are shown in parentheses. The effect of prednisolone was evaluated using Mixed Model 735
Linear Repeated Measures ANOVA and post-hoc Sidak t-test, with dam as subject and litter 736
size as covariate (#P = 0.057; *P < 0.001). Fetal weight (D), placental weight (E) and
737
fetal:placental weight ratio (F) are also shown according to litter size category. The 738
correlation coefficient (r) between total litter size and fetal weight, placental weight and 739
fetal:placental weight ratio for both prednisolone and control data is shown, with P value in 740
parentheses. The effect of prednisolone on fetal and placental weight within litter size 741
categories was analysed by Sidak t-test (P < 0.05). 742
743
Figure 6. The effects of peri-implantation prednisolone treatment on time of delivery
744
and progeny weights. Parturition day was noted to the nearest 12 hours (A). Pups were
745
weighed on day 1 and day 8 after birth (B). Progeny were sexed at 21 days postpartum and 746
weighed every two weeks until week 20 (C, D). Numbers of mated mice are shown in 747
parentheses. The effect of prednisolone was evaluated using Mixed Model Linear Repeated 748
Measures ANOVA and post-hoc Sidak t-test to compare the prednisolone group with the 749
placebo group (mean ± SEM) (#P < 0.10, *P < 0.05 and **P < 0.01).
750 751
752 753