Serum Cytokine Reactions during Pregnancy in Healthy Mares

Abstract

The aim of the research was to determine the differences in IL-1ß, IL-2, IL-4, IL-10, IL-17, INF-γ, TNF-α and IL-1ra blood concentrations in pregnant and non-pregnant healthy mares. The mares were divided into two groups: pregnant mares (n = 18; age 6.11 ± 2.25 years); non-pregnant mares (n = 6; age 5.67 ± 1.75 years). Blood samples were collected from all mares (n = 24) three times every month, and then the pregnant group was divided into three subgroups based on the age of the pregnancy (1st group (n = 6)—6th, 7th, 8th; 2nd (n = 6)—7th, 8th, and 9th; 3rd (n = 6)—8th, 9th, 10th month of pregnancy). The concentrations of IL-1ß, IL-2, IL-4, IL-10, IL-17, INF-γ, and TNF-α were higher in the pregnant than in non-pregnant group, whereas only the concentration of IL-1ra was lower in the pregnant than in the non-pregnant groups. Serum concentrations of proinflammatory cytokines such as IL-1 β, IL-2, and IFN-γ in pregnant mares were greater than in non-pregnant mares at the start of the trimester, whereas, IL-4, IL-17 and TNF-α started to rise in the latter months of the third trimester. IL-10 and IL-1ra concentrations started to decrease after the 6th month of pregnancy. In non-pregnant mares cytokine levels were stable during the whole study. In mares, the change in the ratio between Th1 (IL-1, IL-2, INF-γ, TNF-α) and Th2 (e.g., IL-4, IL-10, IL-13) cytokines occurs during pregnancy. The cytokine profile may be useful in the future for monitoring healthy pregnancies; however, more research is needed especially for miscarriage in mares.

1. Introduction

In mammalian groups of animals, maternal–fetal immunotolerance is mostly regulated by cytokines [1,2]. They function as part of a complex network where the stability between various cytokines aids in the success of reproductive activities [1]. There is two-way communication between the fetus and the mother throughout implantation, and that communication is defined by immunological, endocrine, and molecular variables. Lack of immunotolerance during pregnancy may result in embryo rejection, preterm birth, and other problems for both mother and child [2].

The placenta has been shown to contain a significant number of cytokines [3]. Fetal survival and development depend on this constant conversation between fetal and maternal cells, which is generated by cytokine production and activity on cells at the materno–fetal interface. Pregnancy is frequently referred to as a “Th2 phenomenon” because of the change in the ratio between Th1 (e.g., IL-1, IL-2, INF-γ, TNF-α) and Th2 (e.g., IL-4, IL-10, IL-13) cytokines that occurs during pregnancy [3]. The shift is made to Th2 cytokines, which provide humoral immunity, rather than Th1 cytokines, which are involved in cellular immunity. As a possible tactic for fetal survival in the uterus, the dominance of Th2-type immunity in the Th1/Th2 ratio has been hypothesized [4].

The balance between Th1 and Th2 cytokines is especially important during pregnancy. Immune-inflammatory alterations favoring a Th1 over a Th2 immunological response are commonly observed during the peri-implantation phase [5]. As some studies in mice have shown, abortion can be induced by the administration of high doses of Th1-type cytokines (for example, IL-2), or through the activation of toll-like receptors, both of which lead to the production of Th1-type cytokines [3]. The invasive trophoblast cells gain from, rather than suffer from, the carefully modulated Th1 shift. Th2 immunity, which is prevalent at the placental implantation site, shields the fetus by counteracting the effects of Th1 immunity and allowing for normal fetal and placental growth [5]. In human studies, for example, in comparison to the healthy control group, women with recurrent pregnancy loss (RPL) had a higher Th1 to Th2 cell ratio [6]. What is more, enhanced Th2 immunity throughout pregnancy may trigger autoimmune illnesses or worsen preexisting autoimmune conditions.

So far, not much research has been conducted on the maternal body’s response to the embryo and changes in immune responses in mares. In mares, tissue decomposition during pregnancy is required for implantation, placentation, labor, and delivery [7]. There are several aspects which suggest that horses may be a good model for reproductive studies in humans. The nonsurgical recovery of early-stage embryos and conceptuses, as well as the isolation of pure trophoblast cell populations, are possible because of the horse conceptus’ unique structure and physiology, and such an outcome is not possible in other model organisms [8,9,10,11,12]. However, it appears that horse and human pregnancies are similar in many aspects; more study is needed.

Thus, the aim of this research was to determine the differences in IL-1ß, IL-2, IL-4, IL-10, IL-17, INF-γ, TNF-α and IL-1ra blood concentrations in pregnant and non-pregnant mares as well as the changes of cytokine levels during the third trimester.

2. Materials and Methods

2.1. Animals

The research was conducted in the Polish state stud farm, Dobrzyniewo, which is engaged in conservative breeding of horses of the Polish Konik breed. Out of a herd of 90 Konik Polski horses, 24 Konik Polski mares were selected for blood sampling. All studied mares were housed under the same conditions in all-day open stables. They were fed twice daily with an individualized portion of hay to maintain a healthy condition and had over 12 h of daily access to a large grassy pasture.

Mares were divided based on the ultrasonographic examination of the reproductive tract into one of two distinct groups of mares: non-pregnant and pregnant. The pregnant group’s inclusion criteria were mares that had naturally mated in February and/or March and had a confirmed ultrasonographical pregnancy screened at 14 and 35-days post-ovulation, according to McCue’s protocol [13]. The pregnant group was composed of 18 non-lactating mares (n = 18; age 6.11 ± 2.25 years). The non-pregnant group’s inclusion criteria were lack of mating during the current reproductive season and two negative results of ultrasonographical pregnancy examination. The non-pregnant group was composed of 6 non-lactating mares (n = 6; age 5.67 ± 1.75 years). Ultrasonographic examination of the reproductive tract was conducted using an ultrasound scanner (MyLabOne, ESAOTE, Florence, Italy) and a linear 5 MHz transducer (ESAOTE, Florence, Italy).

The 18 mares enrolled in the pregnant study group were divided into three subgroups based on the age of the pregnancy, where each subgroup included 6 mares. The first subgroup (n = 6) was examined in the 6th, 7th, and 8th month of pregnancy, the second (n = 6) in the 7th, 8th, and 9th month of pregnancy, whereas the third (n = 6) in the 8th, 9th, and 10th month of pregnancy.

2.2. Blood Sampling

Blood samples were collected from all mares (n = 24) three times within one month in October, November, and December. The experimental protocol was approved by the II Local Ethical Committee on Animal Testing in Warsaw on behalf of the National Ethical Committees on Animal Testing (No WAW2/007/2020, 15 January 2020). Peripheral blood was gathered from the jugular vein into sterile K2-ethylenediaminetetraacetic acid (K2-EDTA) tubes for hematological tests and sterile dry tubes for serum analyses using the BD Vacutainer system (BD, USA). The tubes were centrifuged (3000× g, 15 min) and the serum was isolated and stored at −80 C for further analyses.

2.3. ELISA

The concentration of cytokines (IL-1ß, IL-2, IL-4, IL-10, IL-17, INF-γ, TNF-α and IL-1ra) was determined by an available immunoenzymatic commercial assay dedicated to and validated for equine species (Bioassay Technology Laboratory, Shanghai Korain Biotech Co. Ltd., Shanghai, China). Absorbance was measured using a Multiscan Reader (Labsystem, Helsinki, Finland) using Genesis V 3.00 software program.

2.4. Statistical Analysis

Statistical analysis was performed using GraphPad Prism6 software (GraphPad Software Inc., San Diego, CA, USA). Data of blood concentration of eight interleukins/cytokines were presented as data series where each horse represented one realization. Data analysis was performed in the following two steps: (i) testing the differences between data series between non-pregnant and pregnant groups, and (ii) testing the differences between the non-pregnant group and subgroups representing the consecutive months of pregnancy.

In the first step, data series were grouped into non-pregnant and pregnant groups and tested for univariate distributions using a Shapiro-Wilk normality test, independently for each interleukin/cytokine. Comparisons between data series for these two groups were assessed using the unpaired t-test for Gaussian data and the Mann–Whitney test for non-Gaussian data. The alpha value was established as α = 0.05. The numerical data on figures are presented as mean + SD.

In the second step, pregnant mare data series were grouped into three subgroups as follows: (i) 6th, 7th, and 8th month of pregnancy; (ii) 7th, 8th, and 9th month of pregnancy; and (iii) 8th, 9th, and 10th month of pregnancy. Afterward, data series were pooled into five sub-subgroups as follows: 6th; 7th; 8th; 9th; and 10th month of pregnancy. The obtained data series were tested for univariate distributions using a Shapiro–Wilk normality test, independently for each interleukin/cytokine. The comparisons between consecutive months of pregnancy for data series of these five sub-subgroups and the data series of non-pregnant group obtained at the same time laps were assessed using the ordinary one-way ANOVA followed by Tukey’s multiple comparisons test for Gaussian data and the Kruskal–Wallis test followed by the Dunn’s multiple comparisons test for non-Gaussian data. Then, comparisons between non-pregnant and pregnant mares within the sub-subgroups represented by data series for each month of pregnancy (6th, 7th, 8th, 9th, and 10th) and corresponding non-pregnant mares were assessed using the unpaired t-test for Gaussian data and the Mann–Whitney test for non-Gaussian data. The alpha value was established as α = 0.05. The numerical data on figures are presented as mean + SD.

3. Results

Concerning the two main studied groups, the concentration of all eight studied cytokines differed depending on the presence of a pregnancy. The concentrations of IL-1ß, IL-2, IL-4, IL-10, IL-17, INF-γ, and TNF-α were higher in the pregnant than in non-pregnant group, whereas only the concentration of IL-1ra was lower in the pregnant than in non-pregnant group (Figure 1).

Considering months of pregnancy in pregnant mares, specific differences were noted for the consecutive cytokines (Figure 2). No differences (p < 0.05) were found between the concentrations of IL-1ß, IL-2, IL-4, IL-17, INF-γ, and TNF-α measured in the studied months of pregnancy in the pregnant group and the corresponding months in non-pregnant mares. Differences were noted between the concentrations of IL-1ra and IL-10 measured in the studied months of pregnancy in the pregnant group rather than in the corresponding months in non-pregnant mares. In the pregnant group, the concentration of IL-1ra was higher (p = 0.0004) in the 6th than in the 9th and 10th months with no differences between the 6th, 7th, and 8th months as well as the 7th, 8th, 9th, and 10th months (Figure 2B). In addition, in the pregnant group, the concentration of IL-10 was higher (p = 0.0003) in the 6th than in the 8th, 9th and 10th months with no differences between the 6th and 7th months as well as the 7th, 8th, 9th, and 10th months (Figure 2E).

Comparing the two main studied groups specifically, the concentration of IL-1ß was higher in the pregnant than in non-pregnant group in the 7th (p = 0.02), 8th (p = 0.0003), and 9th (p = 0.003) months of pregnancy with no differences in other studied months (Figure 2A). The concentration of IL-1ra was lower in the pregnant than in non-pregnant group only in the 9th (p = 0.0003) and 10th (p = 0.002) months of pregnancy (Figure 2B). The concentration of IL-2 was higher in the pregnant than in non-pregnant group in the 6th (p = 0.03) and 7th (p = 0.04) months of pregnancy with no differences in other studied months (Figure 2C). The concentrations of IL-4, IL-17, and TNF-α were higher in the pregnant than in non-pregnant group in the 8th (IL-4, p = 0.006; IL-17, p = 0.006; TNF-α, p = 0.04), 9th (IL-4, p = 0.05; IL-17, p = 0.05; TNF-α, p = 0.0006), and 10th (IL-4, p = 0.04; IL-17, p = 0.04; TNF-α, p = 0.01) months of pregnancy with no differences in other studied months (Figure 2D, 2F, 2H). The concentration of IL-10 was higher in the pregnant than in non-pregnant group in the 6th (p = 0.0002) month of pregnancy and lower in the 9th (p = 0.0001) and 10th (p = 0.0002) months of pregnancy with no differences in the 7th and 8th months (Figure 2E). The concentration of INF-γ was higher in the pregnant than in non-pregnant group in the 6th (p = 0.02), 7th (p = 0.0007), and 8th (p = 0.01) months of pregnancy with no differences in other studied months (Figure 2G).

4. Discussion

During pregnancy in healthy women, a transition from predominantly Th1-type cytokines to Th2-type cytokines has been noted in peripheral blood mononuclear cell (PBMC) samples, compared to women who were not pregnant, which occurs between the first and third trimesters of pregnancy [14]. The concentrations of IL-1ß, IL-2, IL-4, IL-10, IL-17, INF-γ, and TNF-α were higher in the pregnant than in non-pregnant group, whereas only the concentration of IL-1ra was lower in the pregnant than in non-pregnant group. This phenomenon may relate to the fact that the mare’s body is synthesizing more proteins while being in a pregnancy state. As proteins in the blood help to maintain intravascular volume, during the second and third trimesters of a healthy pregnancy, women’s protein production rises by 15% and 25%, accordingly [14]. In humans uncomplicated preeclampsia patients have considerably decreased blood-protein levels throughout the second trimester of gestation, even before clinical signs of preeclampsia began [14,15]. All the mares were in good health, and there were no miscarriages throughout this time.

In our study serum concentration of proinflammatory cytokines such as IL-1 β, IL-2, and IFN-γ (Th1 response) in pregnant mares were greater than in non-pregnant mares at the start of the trimester. Whereas IL-4 (Th2 response) and TNF-α started to rise in the latter months of the third trimester. In non-pregnant mares the cytokine level was stable during the whole study. As mentioned earlier, pregnancy is frequently referred to as a “Th2 phenomenon” because of the change in the ratio between Th1 and Th2 cytokines that occurs during the course of pregnancy in humans [3] and now we have confirmed that the same reaction occurs in horses. Th2 response also counteracts the Th1 response that allow for Th2 responsiveness to IL-4, for which the serum concentration starts to rise from the 6th month of pregnancy. IL-4 also suppresses Th17 cell responses; however, during the last month of pregnancy the IL-17 and TNF-α serum concentrations started to rise which may relate to low concentrations of IL-10 and IL-1ra–cytokines with strong anti-inflammatory properties.

4.1. IL-1 β, IL-2, and IFN-γ

The effects of proinflammatory cytokines such as IL-1β, IL-2 and IFN-γ on the female reproductive tract and gestation have been extensively studied [16,17,18]. Due to current studies, IL-1β, IL-2, and IFN-γ are detectable in maternal serum at low levels during a healthy pregnancy [19]. Human studies show a considerable drop in IL-1β, IL-2, and IFN-γ levels between the first and third trimesters in healthy women’s pregnancies [17,18,20,21]. In addition, there are some studies demonstrating that these cytokines in maternal blood levels do not fluctuate significantly during a healthy pregnancy [22,23]. Contradictory results may be explained by the inability to detect substantial changes in cytokine concentrations because of their short half-life in the circulatory system [24].

Only a few studies have looked at the relationship between IL-1β serum and endometrial levels in mares. It has been discovered that IL-1β mRNA is present in equine cumulus–oocyte complexes and granulosa cells, whereas horse follicular fluids contain the immunoreactive form of the IL-1β molecule [25]. In one study, proinflammatory cytokine concentrations such as IL-2, TNF-α or IFN-γ rose in the blood in reaction to an experimental paradigm used to induce ascending placentitis [7], which suggests that proinflammatory cytokines play a role in miscarriage in mares as well. Interestingly, ruminants and horses, which have epitheliochorial placentation, do not synthesize IFN-γ in their trophoblast [26], which indicates that trophoblast IFN-γ production is not a universal characteristic. By encouraging endometrial vascular remodeling and angiogenesis at the maternal–fetal border, IFN-γ plays a crucial role in the establishment and upkeep of a successful pregnancy [26], which may explain high levels of IFN-γ at the beginning of the 6th month of pregnancy.

4.2. TNF-α and IL-17

We suggest that increases in the concentrations of both these cytokines may relate to labor preparation in mares. It has been documented that TNF-α levels rose from 108.00 pg/mL to 172.89 pg/mL from the first to third trimesters of normal gestation in women [21]. Overall, it has been suggested in most of the human research that TNF-α rises during pregnancy [13]. In our study, the increase of blood TNF-α level was elevated since the 8th month of pregnancy, which may relate to labor preparation. The pro-abortogenic effects of TNF-α have been linked to a variety of mechanisms such as premature membrane breakdown that is increased by several factors, including trophoblast invasion and placentation and the promotion of pro-apoptotic gene expression in human fetal membranes [27]. In horses, TNF-α is secreted by the endometrium and invasive and non-invasive trophoblast cells from day 30 to 55 of gestation, and it is suggested that it has a role in controlling the placental formation and maternal leukocyte reactions to trophoblasts [28]. In addition, macrophages, most probably by secreting TNF-α, have been linked to local inflammatory mechanisms and tissue reorganization in the uterus during the parturition of mares [29]. Following that discovery, it may be suggested that TNF-α overproduction marks the start of the preparation for delivery in horses as well.

When Th17 cells produce IL-17 cytokine, they trigger the expression of further proinflammatory cytokines, but autoimmune disorders can arise when this process is allowed to run its course unregulated [30]. The fetus is like a semiallogenic graft to the mother’s organism during pregnancy. In our study, it is highly possible that the increase of IL-17 levels in mares’ blood in the latter months of pregnancy is linked to the impending birth. IL-17 concentrations climbed significantly during pregnancy, according to human studies [23]. However, little is known about how it affects natural parturition. Furthermore, the significance of IL-17 in horse parturition has not yet been established [29].

4.3. IL-4, IL-10 and IL-1ra

A fetus is not generally rejected during gestation because of the Th-2 response which would not be possible without IL-4, which plays a critical role in the process [31]. Th2 cell maturation is controlled by IL-4, which also induces the production of inducible Tregs by naïve CD4+ T cells [32]. Aside from Th2 cell activity, IL-4 also influences Treg cell functions, which are critical for a healthy pregnancy [32]. In humans after the third trimester in normal pregnancies, IL-4 concentrations tend to rise [17,20] which is in line with our findings in which the IL-4 level started to rise since the 6th month of a mares’ pregnancy.

In addition, there were changes in concentrations of cytokines with strong anti-inflammatory action such as IL-10 and IL-1ra [30]. IL-10 has a few important positive roles in a healthy pregnancy: it helps with the trophoblast invasion process; aids in placentation; limits inflammation; and controls the functioning of blood vessels; and it was hypothesized that IL-10 amplification during pregnancy also serves as a tolerance technique for the fetal allograft [33]. In our study, very high IL-10 blood concentrations were detected at the 6th month of pregnancy, which may be the consequence of creation the immunotolerance at the beginning of pregnancy. Then as the mares got closer to delivery, the IL-10 blood concentration started to decrease. This is probably related to preparation for a successful delivery. Fetal and placental viability do not depend entirely on IL-10 because mice lacking IL-10 developed abnormal pregnancies but otherwise healthy offspring [34].

In addition, IL-1Ra blood levels were also higher in the early stages of pregnancy in the mares. The concentration level of IL-1Ra as well as IL-10 decreased as parturition approached. It can be seen that first the concentration of IL-10 began to fall, and then IL-1ra. This process can be explained by the fact that IL-10 regulates the amount of IL-1Ra [35].

4.4. Cytokines and Reproductive Problems–Future Directions and Limitations

The decreased expression of anti-inflammatory cytokines such as IL-10, IL-4, IL-1ra in the second trimester may have contributed to the development of preeclampsia, according to some research in humans [13,36]. In addition, many studies have found an association between preterm delivery and an increase in pro-inflammatory cytokines such as IL-1, IL-2, TNF-α [37,38,39]. Increased levels of pro-inflammatory cytokines are linked to preeclampsia, whereas levels of anti-inflammatory cytokines including IL-4 and IL-10 were found to be reduced in humans [32]. In addition, subclinical or asymptomatic illnesses may be detected by measuring cytokine concentrations [13,21].

In horses, there are only a few studies concerned with cytokine changes during miscarriage [7,9]. In our study all the mares were in good health, and there were no miscarriages throughout the study period and no pregnancy problems occurred. Thus, one limitation is connected to having no comparison with problematic mares and future research should focus on measuring changes in cytokine concentrations during pregnancy problems. The other limitation is the small number of animals enrolled in the study. However, the number of animals used was kept to a minimum in order to comply with animal ethics and welfare guidelines. The blood sampling and resining may be stressful especially for pregnant mares leading to miscarriage. Thus, we decided to decrease the number of animals to the necessary minimum. In addition, it would be more convenient to monitor the mares from the beginning of the pregnancy. However, in our case that was impossible because the mares are in the stable only during the late autumn and winter months.

5. Conclusions

During our study we confirmed that the “Th2 phenomenon” also occurs in mares as well as in humans during pregnancy. It is characterized by a change in the ratio between Th1 (IL-1β, IL-2, IFN-γ) and Th2 (IL-4) cytokines that occurs during the course of pregnancy. In addition, an anti-inflammatory state characterized by high levels of IL-10 and IL-1ra was present at the beginning of the third trimester. At the last month of pregnancy, the proinflammatory cytokine concentrations such as IL-17 and TNF-α were elevated, preparing the mare for labor. Thus, monitoring cytokine changes during pregnancy would provide a wealth of useful information. As a result, it would deepen our understanding of how horses react when carrying a foal. Therefore, it may be useful to distinguish between a normal and a problematic pregnancy by studying the changes in the maternal cytokine profile in considerable detail.