Effect of prostaglandin synthesis inhibitors on spontaneous and endotoxin-induced abortion in mice

Ecology of danger-dependent cytokine-boosted spontaneous abortion in the CBA x DBA/2 mouse model: II. Fecal LPS levels in colonies with different basal abortion rates

PROBLEM

Previous data have shown that 'danger' signals, such as bacterial lipopolysaccharide (LPS) acting via toll-like receptors (TLR), are conditions antecedent to early pregnancy failure in several murine abortion models. Indeed, the abortion rate increased in the CBA x DBA/2 model after injection of tumor necrosis factor-alpha (TNF-alpha) + interferon (IFN-gamma) on gestation day (GD) 7.5 only if the LPS-TLR signaling pathway was intact. High rates of cytokine-boosted abortion >80% loss can be demonstrable in certain animal colonies that have a high endogenous (spontaneous) rate of resorption (30-50%). A specific role for LPS on GD 0.5 determines the endogenous loss rate and on GD 6.5 the responsiveness to cytokine boosting of losses. Th1 cytokines (or the stress that induces these cytokines) increase intestinal permeability and absorption of luminal contents. It was predicted that intestinal availability of LPS was a major factor in the endogenous and cytokine-boosted resorption rates.

METHOD OF STUDY

A fixed weight of fresh mouse droppings from CBA/J female mice house in a high-abortion-rate (30-40%) colony at Clamart, France was homogenized and filtered. Fresh mouse droppings from a low-abortion-rate (10-15%) colony in Berlin, Germany were similarly processed. LPS was assayed using the Limulus amoebocyte lysate bioassay.

RESULTS

To our surprise, there was no significant difference in LPS content of fecal samples from the two colonies.

CONCLUSION

A high endogenous rate of abortion and cytokine (or stress) boosted abortion in the CBA x DBA/2 model is not explained by the LPS content of feces. Possible explanations include: fecal LPS does not reflect small intestinal LPS, there are additional TLR signals besides LPS that are important and endogenous stress levels may be higher in high-abortion-rate colonies, so permeability of the intestine (and Th1 cytokine levels) may be already higher. These data have implications for studies on the role of flora in human pregnancy problems.

Cytokines: Important for implantation?

PROBLEM

Cytokines are obviously very important in an established pregnancy, but what about human embryo implantation?

METHODS

Literature review.

RESULTS

We first discuss the necessity and limits of animal models, and then review the few cytokines which have been demonstrated by knock-out methods to be absolutely necessary for embryo implantation using in animal models. We then review what is known or discussed about the role of other cytokines as deduced from quantitative and/or qualitative dysregulation in animals and in humans.

CONCLUSIONS

Cytokines are indeed involved in implantation as they are in ongoing pregnancy and delivery. Relevance to infertility and recurrent pregnancy loss is discussed.

Toll-like receptor 4 polymorphisms and idiopathic chromosomally normal miscarriage

BACKGROUND

Lipopolysaccharide (LPS or endotoxin) exposure resulting from microbial invasion of the endometrium disturbs the Th1/Th2 balance at the feto-maternal interface and has been linked to the risk of idiopathic miscarriage in a range of human and animal studies. Toll-like receptor 4 (TLR4) mediates LPS signalling, and the human TLR4 gene harbours two single-nucleotide polymorphisms (SNPs) known to reduce LPS responsiveness. We hypothesized that genetic variation altering TLR4 function may influence the risk of idiopathic pregnancy loss.

METHODS AND RESULTS

We examined fetal TLR4 genotypes in a case-control cohort of chromosomally normal miscarriages (n=96) and healthy term newborns (n=113). The allele frequencies of the Asp299Gly and Thr399Ile TLR4 SNPs were determined by quantitative PCR using DNA extracted from extraembryonic tissues and umbilical cord blood, respectively. TLR4 genotype frequencies were not significantly different between cases and controls.

CONCLUSIONS

There was no association between fetal TLR4 polymorphisms, Asp299Gly and Thr399Ile, known to blunt LPS responsiveness, and the risk of idiopathic, chromosomally normal miscarriage. Nevertheless, TLR4 or perhaps other LPS-binding chaperone molecules are biologically plausible candidate genes that may alter the risk of idiopathic miscarriage.

Anti-inflammatory pharmacotherapy during pregnancy

NSAIDs or cyclooxygenase inhibitors (COX inhibitors), including aspirin, are widely used to treat pain, fever and the articular symptoms of chronic rheumatic diseases. Manifestations of connective tissue or autoimmune diseases are commonly treated with glucocorticosteroids. The effect and side effects of NSAIDs depend on the isoforms of cyclooxygenases that they preferentially or selectively inhibit. The use of COX inhibitors has recently been associated with infertility and miscarriage. The classical nonselective COX inhibitors, including aspirin, do not increase the risk of congenital malformations in humans but administered in the latter part of gestation, they can affect pregnancy and the fetus. The ability of nonselective and selective COX inhibitors to prolong gestation has been used by obstetricians to inhibit premature delivery. The vascular effects of prostaglandin inhibitors can cause constriction of the fetal ductus arteriosus and reduce renal blood flow. These complications have been described for most nonselective COX inhibitors but are increasingly reported also for the selective COX-2 inhibitors. Aspirin, which causes irreversible inhibition of cyclooxygenases, differs from other NSAIDs with regard to indication, effects and side effects. Prematurity, which is increased in pregnancies of women with connective tissue diseases, is an additional risk factor for adverse effects of antenatal exposure to NSAIDs. Therefore, treatment with COX inhibitors should be discontinued at week 32 of gestation. The ability of NSAIDs to compromise reproductive function by inhibition of ovulation and as causative agents for miscarriage is still under debate. Glucocorticosteroids given in early pregnancy are a risk factor for the development of oral clefts. Therefore, the daily dose should be kept to <or= 15 mg during the first trimester. High doses of glucocorticosteroids in the second and third trimester are reserved for flares of autoimmune diseases. Intrauterine fetal growth restriction and premature delivery are possible side effects of high doses.

Ecology of Danger-dependent Cytokine-boosted Spontaneous Abortion in the CBA × DBA/2 Mouse Model. I. Synergistic Effect of LPS and (TNF-α + IFN-γ) on Pregnancy Loss

PROBLEM

Previous data have shown "danger" signals, such as bacterial lipopolysaccharide (LPS) acting via toll-like (tlr) receptors are required for early pregnancy failure in several murine abortion models. Indeed, the abortion rate increased in the CBA x DBA/2 model after a gestation day (gd) 7.5 injection of tumour necrosis factor (TNF)-alpha + interferon (IFN)-gamma only if the LPS-tlr signalling pathway was intact. High rates of cytokine-boosted abortion >80% loss can be achieved in certain animal colonies, that have a high endogenous (spontaneous) rate of resorption (30-50%). A specific role for LPS has been postulated to determine both the endogenous and cytokine-boosted losses.

METHODS

To test the role of LPS in spontaneous and cytokine-boosted abortions, recombinant TNF-alpha + IFN-gamma, and LPS were injected in different doses and sequences intraperitoneally (i.p.) into CBA x DBA/2 mated mice in the Toronto General Research Institute animal facility where the endogenous abortion rate is <30%. The effects of poly IC, a tlr3 agonist that induces IFN-gamma that can reverse LPS-induced tolerance, and effects of anti-MD-1 on TNF-alpha induction by LPS, poly IC, CPG, or HSP in vitro were also examined.

RESULTS

A high endogenous rate of loss similar to that seen in Clamart could be achieved by increasing exposure to LPS on the morning after mating (gd 0.5). The magnitude by which the abortion rate could be increased by an i.p. injection of 2000 u TNF-alpha + 1000 u IFN-gamma on gd 7.5 was independent of the endogenous rate of loss, and could not be increased by doubling the dose. One microgram of LPS given on day 7.5 achieved a similar rate of loss, and if given with the cytokines, synergistically boosted the rate of loss to near Clamart rates. LPS given 1 day prior to the cytokines abrogated the cytokine effect, whereas LPS given day 0.5 had no significant effect on the response to day 7.5 cytokine injection. Blocking MD-1 inhibited TNF-alpha stimulation by poly IC, LPS, CPG, or HSP in vitro, and reduced abortion rates. Poly IC did not avert LPS-type tolerance effects in vivo.

CONCLUSIONS

High endogenous rates of abortion in the CBA x DBA/2 model may be explained by exposure to LPS at the time of mating. Increased rates of loss triggered by cytokines later in pregnancy may depend on increased absorption of LPS from intestinal flora.

Searching for links between endotoxin exposure and pregnancy loss: CD14 polymorphism in idiopathic recurrent miscarriage

PROBLEM

Lipopolysaccharide (LPS) (endotoxin) is a well-known inducer of abortions in mice. In addition it has been proposed that gut-derived LPS of gram-negative bacteria may play a role in triggering idiopathic recurrent miscarriage (IRM) in humans. CD14 is one of the key molecules that mediates the effects of LPS. Promoter region polymorphism (-159C/T) in the CD14 gene is functionally important by regulating CD14 levels. High-producing CD14 genotype (TT) associates with deleterious effects of gut-derived LPS in hepatic cirrhosis in humans. It is not known whether women with IRM are genetically more prone to suffer from toxic effects of LPS.

METHOD OF STUDY

By using polymerase chain reaction we analyzed the CD14 promoter region polymorphism in 38 women with IRM and in 127 normal controls of Finnish origin.

RESULTS

There were no significant differences in the CD14 (-159C/T) allele or the genotype frequencies between the IRM women and the controls. However, there was a trend associating the presence of the T allele with increased odds of miscarriage.

CONCLUSIONS

Although we were not able to find a statistically significant association between CD14 genotypes and IRM in our relatively small study population, a further study with a larger sample size is warranted to explore the role of high-producing CD14 genotypes in IRM. Also studies highlighting environmental LPS triggers and other intrinsic mediators of LPS signalling are needed to solve the enigmatic role of LPS in IRM in humans.

Effects of endotoxin and macrophage-related cytokines on the contractile activity of the gravid murine uterus

Immune cell trafficking and activity are implicated in the parturition process, but little is known about the role of macrophages in control of uterine contractility at term. In the present study, we tested the hypothesis that endotoxin (lipopolysaccharide [LPS]) enhances uterine contractile activity through a mechanism that involves activation of resident macrophages. Various uterotonins and anti-inflammatory mediators were added to a standard muscle bath preparation that contained strips of uterus from Day 15 pregnant C3H/HeN mice. Spontaneous and agonist-induced contractile activity was enhanced following LPS treatment. LPS increased amplitude but not frequency of contractions. Addition of anti-inflammatory cytokines, interleukin 10 or transforming growth factor beta, to suppress macrophage activation did not block LPS-induced increases in contractility. By contrast, indomethacin given to block prostaglandin production suppressed the LPS-induced increase in amplitude of contractions. These findings suggest that an inflammatory response, possibly mediated by activation of macrophages and prostaglandins, participates in the regulation of amplitude but not frequency of contractile activity by the murine uterus before onset of parturition.

MD-1 is a critical part of the mechanism causing Th1-cytokine-triggered murine fetal loss syndrome

PROBLEM

Fetal loss syndrome (abortion/resorption) occurring on or after gestation day (gd) 9.5 in CBA/JxDBA/2 matings is dependent upon presence of TNF-alpha + IFN-gamma, which act by increasing expression of fg12 prothrombinase at the feto-maternal interface. The magnitude by which the abortion rate can be boosted by an injection of these cytokines on gd 7.5 depends on endogenous rate of loss, and appears to depend on microbial flora. Is cytokine-triggered abortion dependent upon a third signaling pathway that senses 'danger'?

METHODS

Female CBA/J were mated to DBA/2 males and, C57B1/6 and C57B1/6 TNFalphaR1-/-Mak were mated to C57B1/6 control or TNFalphaR1-/-Mak males. LPS from Escherichia coli and Salmonella enteritidis, or the combination of TNF-alpha + IFN-gamma, was injected to stimulate abortions. The effect of anti-MD-1, which interferes with expression of CD14 and, hence, with signaling by LPS via the CD14-tlr4 complex, on TNF-alpha + IFN-gamma was tested. The presence of MD-1 in the uterus was evaluated by in situ hybridization, and effect of lipopolysaccharide (LPS) on mice lacking TNF-alphaR1 was tested.

RESULTS

Anti-MD-1 completely abrogated TNF-alpha + IFN-gamma-induced abortions. MD-1 was expressed on trophoblast and in deciduas on gd 8.5 but LPS could not abort mice that lacked the type 1 receptor for TNF-alpha. Pregnant CBA/J females had classical resorptions (abortions) countable on gd 13.5-14.5 in response to LPS from E. coli or S. enteritidis, but C57B1/6 strain mice resorbed only in response to the latter, and E. coli LPS appeared to induce 'occult' losses. 'Occult' loss did not require TNF-alphaR1.

CONCLUSIONS

TNF-alpha + IFN-gamma could not induce murine abortions without co-presence of a 'danger' signal such as LPS acting via CD14 on toll receptors, and LPS could not act without co-signaling by TNF-alpha. Classical resorptions/abortions and 'occult' losses have a different mechanism in these models as reflected in type of endotoxin and requirement for TNF-alphaR1 signaling.

Anti-abortive effect of Radix scutellariae and Rhizoma atractylodis in mice

The objective of this study is to investigate the significance of natural killer (NK) cells and interleukin-2 in uterus in the early embryo loss (or resorption), and to elucidate the immunological modulation of maternal-fetal interface with Chinese herbal medicine Radix scutellariae (huang qin) and Rhizoma atractylodis (bai zhu). Lipopolysaccharide (LPS) was given via the tail vein to induce abortion in mice at day 7 of gestation. Uterine NK cells and IL-2 contents were analyzed by immunohistochemistry and enzyme linked immunosorbent assay (ELISA), respectively. The number of NK cells was found to be much higher (mean = 180 +/- 39) in the decidua of LPS-treated abortion mice. But when the Chinese herbal medicine was used to prevent LPS-induced abortion, less NK cells (mean = 11 +/- 4) were counted (p < 0.01). The mean value of IL-2 in LPS-treated mice was 5.25 +/- 2.5938 pg/mg protein, higher than (p < 0.05) that of the herb prevention group, which was only 1.86 +/- 0.9789 pg/mg protein. The results therefore indicate that the increase of NK cells in the decidua and IL-2 contents in the uterus in LPS-treated mice is closely related to the embryo loss, and that the Chinese herbal medicine prescription composed of Radix scutellariae and Rhizoma atractylodis has an anti-abortive effect through inhibition of maternal-fetal interface immunity.

Thinking outside the box: mechanisms of environmental selective pressures on the outcome of the materno-fetal relationship

PROBLEM

Study of mechanisms causing spontaneous abortion of the vascularized placenta have focused primarily on the feto-maternal immunological relationship within the pregnant mother. The Th1 cytokines such as tumor necrosis factor (TNF)-alpha + interferon (IFN)-gamma derived in part from natural killer (NK) and NKgammadeltaT cells have been implicated in causing abortion via up-regulation of the novel prothrombinase fgl2 at the feto-maternal interface; Th2/3 cytokines such as interleukin (IL)-10, progesterone-induced blocking factor (PIBF), and TGF-beta2 derived from gammadeltaT cells stimulated by embryo antigens in the context of the OX-2 (CD200) tolerance signal have been viewed as counteracting the Th1 effect. These mechanisms are distinct from those causing and preventing occult pregnancy loss during the periimplantation phase of pregnancy prior to development of a vascularized placenta. Spontaneous abortions in the CBA/J x DBA/2 can be boosted by injecting TNF-alpha + IFN-gamma, but the boosted abortion rates can range from < or = 30 to > 80%, depending on the loss rate in uninjected mice, and this is not explainable by the endogenous level of these cytokines. Furthermore, there is a poor correlation between Th1/Th2.3 cytokine ratios and abortion rates. Could there be a third factor involved, and if so, what might this mean?

METHODS

Known precipitants of recurrent abortion in mice were reviewed with particular attention to stress and endotoxin absorption. The effect of antagonizing the response to bacterial lipopolysaccharide (LPS) (endotoxin) was tested. Data on environmental selective pressures were considered (i.e. thinking outside the 'box', which typifies the conventional approach to thinking about materno-fetal interactions).

RESULTS

Th1 cytokine-triggered abortions appear to depend on availability/presence of LPS.

CONCLUSIONS

Environmental selective pressures are implicated in eliminating 'genetically weaker' embryos in early pregnancy.

Procoagulants in fetus rejection: the role of the OX-2 (CD200) tolerance signal

The spontaneous loss of normal karyotype embryos may be initiated or prevented by the maternal immune system. In mice, loss between the time of implantation (day 4.5) and formation of a vascularized placenta (day 9.5) when the embryo is too large to survive by diffusion alone, is analogous to occult pregnancy failure in humans. They are called occult because usually the woman does not know she is pregnant. From studies in mice, these early losses have a different mechanism than abortion of a vascularized placenta (analogous to clinically evident human spontaneous miscarriage). The latter depend on the activation of the novel prothrombinase fgl2 on the fetal trophoblast and in maternal decidua by the T helper-1 (Th1) type cytokines TNF- alpha+gamma -interferon that arise from NK cells and NK gammadelta T cells; conversion of prothrombin to thrombin which in turn generates IL8 that activates polymorphonuclear leukocytes leads to embryonic death. These inflammatory processes are counteracted by Th2/3-type cytokines that arise in part from V gamma 1 delta 6 T cells reacting to, as yet, unidentified trophoblast antigens in the presence of the 'tolerance signaling molecule' OX-2. By contrast, peri-implantation losses (between implantation and formation of a vascularized placenta, analogous to occult losses in humans) appear to be dependent upon perforin(+)cells, complement activation, and products of alphabeta T and NK alphabeta T cells, but not on TNF- alpha or procoagulant activation. Similarities and differences between findings in the mouse and human, and the potential evolutionary significance of mechanisms affecting reproductive success are reviewed.

Cyclooxygenase is involved in the effects of progesterone-induced blocking factor on the production of interleukin 12

OBJECTIVE

Immunologic effects of progesterone are mediated by the progesterone-induced blocking factor. Progesterone-induced blocking factor inhibits natural killer cytotoxic activity and arachidonic acid release from mononuclear cells. The relationship between increased prostaglandin synthesis and increased cytotoxic activity of the lymphocytes is still unclear; therefore we investigated the effect of progesterone-induced blocking factor-neutralizing antibody, as well as simultaneous indomethacin treatment, on interleukin 12 production.

STUDY DESIGN

Pregnancy lymphocytes were treated with anti-progesterone-induced blocking factor antibody or lipopolysaccharide as a positive control in the presence or absence of indomethacin. Interleukin 12 production by peripheral blood mononuclear cells was detected by immunocytochemical examination. The 2-tailed Student t test was used for statistical evaluation.

RESULTS

Neutralization of progesterone-induced blocking factor, as well as lipopolysaccharide treatment, resulted in an increased expression of interleukin 12 that was corrected by simultaneous indomethacin treatment.

CONCLUSION

Progesterone-induced blocking factor reduces the expression of interleukin 12 via the inhibition of arachidonic acid metabolism. This results in lowered cytotoxic natural killer activity, which favors a normal pregnancy outcome.

Maturation of spontaneous and agonist-induced uterine contractions in the peripartum mouse uterus

This study tested the hypothesis that the uterus achieves maximum contractile capabilities before the onset of labor. Basal and agonist-stimulated contractions were assessed in uterine strips on Day 15 or 18 of pregnancy, the day of parturition, or 1 day postpartum (n = 4-13 per group). Spontaneous contractions were evident in all groups (n = 4-13 per gestational group); contraction frequency was greater in peripartum groups than in virgin controls ( approximately 4.6 versus 2.8/200 sec). Peak amplitude was nearly 9-fold higher on Days 15 and 18 and over 30-fold higher in the postpartum and 1 day postpartum groups than in nonpregnant mice. Maximum frequency and peak amplitude were achieved in response to 10(-6) to 10(-8) M oxytocin or arginine vasopressin (OT(max) or AVP(max)). Frequency of contractions in response to OT(max) peaked on Day 18 and then declined. Contraction amplitude increased 5-fold on Day 15, declined on the day of birth (equivalent to nonpregnant level), then rebounded to peak on postpartum Day 1. AVP(max) similarly increased frequency and amplitude of contractions, except that maximum contraction amplitude occurred postpartum. Thus, an endogenous oscillator, residing in the uterus, sustains high basal and agonist-induced contraction frequency during pregnancy. Although acceleration of this pacemaker occurred before term, the data suggest that peripartum increases in contraction amplitude characterize the transition to the powerful synchronous contractions of parturition.

Femoral vein ligation increases bone mass in the hindlimb suspended rat

Bone remodeling in response to changing mechanical demands is well recognized. It has been hypothesized that alterations in interstitial fluid flow (IFF), due to intraosseous pressure changes, influence bone remodeling. The goal of this study was to investigate the role of IFF in bone in the absence of mechanical strain using an in vivo model, the hindlimb suspended rat. Bone remodeling was assessed by direct measurements of weight, dimensions, bone mineral content (BMC) and bone mineral density (BMD) by dual-energy X-ray absorptiometry (DEXA), and trabecular density using peripheral computed tomography (pQCT). Ligation of one femoral vein was performed as a means to alter the IFF within the ipsilateral femur; the contralateral limb was sham-operated as control. Animals were suspended for a period of 19 days. Intramedullary pressure in the venous-ligated femurs increased relative to the sham-operated control femurs (27.8 mmHg vs. 16.4 mmHg, p < 0.05), suggesting venous ligation increased IFF proportional to the pressure drop across the bone. Bone mineral content (BMC), when normalized to body weight, increased significantly in the venous-ligated femurs relative to control limbs (115.9 +/- 15.6% vs. 103.8 +/- 13.2%, p < 0.001); similarly, gains in length (106.2 +/- 2.4% vs. 104.5 +/- 2.1%, p < 0.05) and distal width (110.8 +/- 10.3% vs. 106.2 +/- 8.2%, p < 0.05) for the femurs with venous ligation were significantly greater relative to sham control. Furthermore, trabecular density was significantly higher in the femurs with venous ligation (351 +/- 12 g/cm3 vs. 329 +/- 11 g/cm3, p < 0.05). Daily administration of the cyclooxygenase inhibitor, indomethacin, via drinking water, suppressed the length increases observed for the venous ligated femur, suggesting a role for prostaglandins in IFF-mediated remodeling. These results suggest that IFF can directly influence bone adaptation independent of mechanical loading, and supports the hypothesis that fluid flow modulates bone remodeling.

Why did your mother reject you? Immunogenetic determinants of the response to environmental selective pressure expressed at the uterine level

PROBLEM

Maternal "rejection" of the implanted conceptus is considered to account for a significant proportion of miscarriages (abortions) in both humans and animals. Our understanding of mechanisms has been limited, and hence, explanations for nonrejection have remained largely speculative. Losses, when they occur, could represent either random accidental failure of protective mechanisms or a more purposeful discrimination.

METHOD OF STUDY

An analysis of the most recent data.

RESULTS AND CONCLUSIONS

The embryo is most akin to a parasite, and pregnancy is most akin to a host-parasite interaction. If one excludes chromosome abnormalities in the embryo as a cause of death, activation of coagulation mechanisms, leading to vasculitis affecting the maternal blood supply to the implanted embryo, appears to represent a major loss-causing mechanisms--a form of ischemic autoamputation. Proinflammatory T-helper (Th) 1-type cytokines trigger this process via upregulation of a novel prothrombinase, fgl2. Th2/3 cytokines, such as interleukin (IL)-4, IL-10, and transforming growth factor (TGF)-beta 2, may antagonize the processes involved. Cytokine balance is determined by the genetics of the mother, which regulate her response to stress; endotoxin (LPS); and paternal antigens, selectively expressed on the trophoblast of the embryo, via imprinting. Based on studies in abortion-prone mice, where immunity to paternal alloantigens prevents loss, three distinct gene products in the embryo are proposed to determine the cytokine response to maternal lymphomyeloid cells in the uterus.

Role of interferon-gamma in the priming of decidual macrophages for nitric oxide production and early pregnancy loss

We have previously shown that both priming and triggering signals were needed for nitric oxide production by decidual macrophages and that nitric oxide was responsible for embryo wastage. In this study, we investigated the role of IFN-gamma as the primary signal for macrophage activation in early embryo loss. IFN-gamma-deficient (GKO) and heterozygous F1 control mice were injected with lipopolysaccharide (LPS) at day 7 of gestation. The results showed that the GKO mice were more resistant to LPS-induced embryo loss than the wild type. This suggested that IFN-gamma was needed for LPS-induced embryo resorption and that decidual macrophages from pregnant GKO mice were not primed and could not be activated when given LPS. Further, the results showed that IFN-gamma mRNA was simultaneously expressed in the same embryos that also expressed mRNA markers for macrophage activation (TNF-alpha and iNOS), indicating that macrophage activation could be a consequence of IFN-gamma production. Similarly, we investigated the role of IL-12 as a switch cytokine capable of eliciting TH1-associated cytokine production including IFN-gamma. The results showed that IL-12 mRNA expression was correlated with IFN-gamma expression and macrophage activation. In this in vivo study, we showed for the first time that spontaneously increased decidual IFN-gamma expression is detrimental to embryo survival.

Soluble receptors neutralizing TNF-alpha and IL-1 block stress-triggered murine abortion

PROBLEM

In several models of abortion in rodents, the success or failure of the implanted embryos is determined by a balance between pro-inflammatory cytokines such as tumor necrosis factor-alpha (TNF-alpha), interleukin-2, interleukin-1 (IL-1), and gamma-interferon, and cytokines that counteract the former, such as interleukin-10 and transforming growth factor-beta 2 (TGF-beta 2)-related suppressor factor. Stress can trigger abortions in susceptible strains of mice and is thought to reflect the pathogenesis of some types of miscarriage in human pregnancy. In mice, stress increases levels of the abortogenic cytokine TNF-alpha and decreases the suppressive activity of TGF-beta 2-related factor via a neurotransmitter substance P (SP)-dependent pathway. Evidence for a role of pro-inflammatory cytokines in SP-mediated abortions in vivo is indirect.

METHODS

Direct evidence for a role of IL-1 and TNF-alpha in stress-triggered abortions was sought by injecting pregnant female mice with soluble receptors neutralizing TNF-alpha (rhuTNFR:Fc) or IL-1 (rmIL-IR) beginning 1 day after implantation and prior to stress.

RESULTS

The stress-triggered abortion rate was reduced by 68% when either TNF-alpha or IL-1 antagonists were injected. The stress-triggered decreased TGF-beta 2-like suppressive activity in the maternal uterine decidua was not restored by injection of either antagonist; indeed the soluble IL-1 receptor significantly reduced suppressive activity in unstressed control mice, and soluble TNF-alpha receptor had a lesser effect.

CONCLUSIONS

Both IL-1 and TNF-alpha play a role in the pathogenesis of stress-triggered abortions, and may induce a compensatory physiological increase in suppressive activity in normal pregnancy counteracting pro-inflammatory cytokines.

Psycho-neuro-cytokine/endocrine pathways in immunoregulation during pregnancy

PROBLEM

Some mammalian pregnancy failure is thought to occur by immunological or immunologically modifiable mechanisms. The original model wherein spontaneous abortion was proposed to represent rejection of the conceptus as an allograft has been supplanted by a model of maternal paraimmunological natural effector cell toxicity to fetal trophoblast more closely related to tumor rejection. The problem is to integrate current information concerning the role of immunological, paraimmunological, endocrinological, and stress-triggered neural factors that determine whether or not abortion will occur.

METHODS

Review of existing data.

RESULTS

An integrated model is proposed.

CONCLUSION

Immunological factors play an important role in abortion processes and prevention of abortions. The existence of abortogenic mechanisms and their regulation appears to be based upon optimizing survival of the species. Two new conceptual models provide a useful framework for further investigation of human pregnancy failure and its treatment.

Immunological prevention of spontaneous early embryo resorption is mediated by non-specific immunosimulation

PROBLEM

Spontaneous early embryo resorption following implantation occurs in many species, but little is known regarding the causes or the prevention of early pregnancy failure. Embryo and fetal loss have widely been assumed to be due to maternal allospecific immune rejection. Alloimmunization therapy with paternal tissues has been successfully used in human and murine pregnancies to prevent early embryo demise. The mechanisms of this treatment have been assumed to be the induction of antigen specific, fetal "graft" enhancing antibodies and suppressor cells. The purpose of this study was to investigate the validity of this assumption.

METHOD

To investigate these general assumptions, female CBA/J mice were immunized with either specific or nonspecific antigens prior to mating with DBA/2 or Balb/c males. Further, a model system for the study of bacterial lipopolysaccharide (LPS) induced abortion was used to demonstrate the nature of antigen specific immune protection against abortion.

RESULTS

Whereas the administration of 1 microgram of LPS to CFW female x CFW male pregnant mice on day 7 of gestation induced complete fetal resorption, prior immunization with 20 micrograms of LPS completely prevented LPS induced abortion as long as the anti-LPS antibody titers remained above a threshold value of about 1/500. Therefore, early embryo loss could be induced by a bacterial infection and could be prevented by appropriate immunity to abortogenic factors. However, due to the short half-life of IgM antibodies, immunity to LPS was short-lived and the protective effect of LPS immunization against LPS induced abortion waned after 5 wk. Through the use of the CBA/J female x DBA/2 male model system to study spontaneous early embryo loss, previous vaccination of CBA/J female mice with Balb/c spleen cells expressing paternal MHC antigens, complete Freund's adjuvant (CFA) or LPS, all decreased the incidence of spontaneous resorption in subsequent pregnancies. Similarly, a previous mating with a Balb/c male prevented spontaneous embryo loss for a period of up to 6 wk. However, none of the immunotherapeutic vaccinations or matings had a permanent effect on CBA/J female x DBA/2 male embryo survival, which one would have expected if specific immune mediators were involved.

CONCLUSION

The results of this study indicated that the decrease in the incidence of spontaneous embryo resorption following alloimmunization was more likely to be due to nonspecific immunomodulatory effects on the immune system of the female mice, as opposed to specific antipaternal immunity. This may, in part, explain the placebo effects observed for alloimmunization therapy for human habitual pregnancy loss. The relevance of these results to the development of immunotherapy strategies for the prevention of habitual abortion is discussed.

Is spontaneous resorption in the DBA/2-mated CBA/J mouse due to a defect in "seed" or in "soil"?

PROBLEM

Recurrent spontaneous resorption in DBA/2-mated CBA/J mice has been attributed to damage by NK-lineage cells and TNF-alpha beginning several days after implantation. However, some recent data have suggested CBA/J female mice have a high proportion of preimplantation embryo abnormalities resulting in defective in vitro implantation and impaired trophoblast outgrowth. Could spontaneous abortion be due to a defective embryo ("seed") rather than a hostile post-implantation uterine lining ("soil").

METHOD

Mated CBA/J females were manipulated so as to have high spontaneous abortion rates and a high percent abnormal embryos, or low resorption rates and a low percent abnormal embryos. Embryos from low aborting females were transferred into high aborting female recipients that were pseudopregnant, and vice versa.

RESULTS

Abnormal embryos from females with high abortion rates implanted in low aborting females and did not show any greater tendency to resorb than normally developed embryos in these recipients. By contrast, normal embryos to some extent and abnormal embryos to a much greater extent, gave a high abortion rate when the recipient female was a high aborter.

CONCLUSION

Properties of the "soil" into which embryos implant determines the likelihood of success or failure (abortion). Abnormal pre-implantation embryos can be "rescued" by "good soil"; "sick soil" damages both normal and abnormal embryos. Defining the cellular and molecular mechanisms may be useful in understanding basic mechanisms leading to aborting and nonaborting pregnancy.

Intralipid as treatment for recurrent unexplained abortion?

PROBLEM

Safe, effective, and inexpensive alternatives to partner leukocyte immunotherapy are being sought. Psychotherapy may be effective but it is uncertain what constitutes effective treatment and the form of treatment tested in cohort controlled trials is expensive. IVIG also appears effective, but is expensive.

METHOD

A published double blind randomized controlled trial in which Intralipid (Kabi Vitrum, Toronto, Ontario) was used as a control versus trophoblast membrane vesicles was reviewed. A prediction made from this data was then tested using the DBA/2-mated CBA/J mouse model of recurrent spontaneous abortion.

RESULTS

It can be hypothesized from the human clinical trial data that Intralipid even in small doses could be an effective antiabortion treatment. The number of patients in the published study is too small for the required degree of precision. Intralipid was highly effective in preventing abortion in mice, and protection was prolonged. This may be explained by previous data in the literature showing that Intralipid affects the reticuloendothelial system of the recipient.

CONCLUSIONS

The evidence suggests that Intralipid might be an effective treatment for human recurrent miscarriages, and injection into women who may become pregnant has been found ethically acceptable at one university center. Comparison of Intralipid to partner leukocyte immunotherapy or IVIG would be worthwhile. For adequate statistical power, this would require a large, multicenter, prognostically stratified randomized controlled trial and could be accomplished via the Recurrent Miscarriage Immunotherapy Trialists Group network.

Spontaneous abortion in immunodeficient SCID mice

PROBLEM

Immunodeficient SCID mice on the CB-17 have been used to test the role of "rejection" in a xenogeneic blastocyst transfer model of recurrent miscarriage, but interpretation of the data requires knowing syngeneic within-species matings have a high success rate and do not require immunotrophic factors expected only in immunocompetent non-T-cell deficient mice.

METHOD

Resorption rates were studied in a SCID CB-17 barrier facility that provided the mice used to test the role of immunology in the resorption model.

RESULTS

Spontaneous resorption in syngeneically mated immunodeficient SCID mice on the CB-17 background occurred at an unexpectedly high rate and could not be prevented by treatment with anti-asialo GM1 antibody or GM-CSF, both of which are effective in ameliorating abortion in DBA/2J-mated CBA/J mice. Immunocompetent CB-17 +/+ mice showed an even higher rate of loss. The latter was also not affected by treatment with antiasialo GM1 antibody or by GM-CSF and was not prevented by tetracycline (which is effective in the DBA/2-CBA/J system) or progesterone treatment. Mating experiments showed a scid/+ x scid@+ cross gave the highest rate of loss, and it appeared that the presence of +/(+)-type embryos in the uterus could be augmenting abortion with selective discrimination against scid/scid embryos. High abortion rates were associated both with appearance of a coagulase-negative Staphylococcus sp. in feces and with loss of one component of the SPF flora. Decidual tissue from mated CB-17 +/+ mice showed premature release of TNF-alpha in absence of TGF-beta 2-related suppressor activity, and vascular lesions (fibrinoid necrosis), varying in extent, were associated with both scid/scid x scid/scid and +/+ x +/+ pregnancies. TNF-alpha also appeared prematurely in pregnant scid/scid mice, but the levels were lower (and areas of necrosis smaller than in +/+ x +/+ pregnancies). Outcrossing onto a C57B1/6 background dramatically reduced the abortion rate, indicating an important genetic effect on susceptibility with heterogeneity protecting against abortion.

CONCLUSIONS

SCID mice on the CB-17 background do not have a high rate of successful syngeneic pregnancies, and a TNF-alpha induced vasculopathy may be responsible. Abortion was not caused by immunodeficiency leading to loss of immunotrophism because immunocompetent non-SCID CB-17 mice had a higher rate of loss. Factors augmenting the abortion rate included the presence of embryos of the +/+ genotype in the uterus and treatment with anti-asialo GM1 antibody. Abortion rates were not reduced by treatments effective in the DBA/2-mated CBA/J mouse model but were reduced by re-establishing a new colony with defined flora (a temporary effect) and by outcrossing mice with a different (C57B1/6) background. Together, the data suggest an infectious trigger (identity uncertain) of the vasculopathy and an important genetic influence on susceptibility with heterozygosity and a SCID mouse mutation providing against abortion a degree of protection.