The Interface of the Immune and Reproductive Systems in the Ovary: Lessons Learned from the Corpus Luteum of Domestic Animal Models

Authors

  • Joy L. Pate,

    1. Department of Dairy and Animal Science, Center for Reproductive Biology and Health, The Pennsylvania State University, State College, PA, USA
    Search for more papers by this author
  • Koji Toyokawa,

    1. Department of Dairy and Animal Science, Center for Reproductive Biology and Health, The Pennsylvania State University, State College, PA, USA
    Search for more papers by this author
  • Sadhat Walusimbi,

    1. Department of Dairy and Animal Science, Center for Reproductive Biology and Health, The Pennsylvania State University, State College, PA, USA
    Search for more papers by this author
  • Edyta Brzezicka

    1. Department of Dairy and Animal Science, Center for Reproductive Biology and Health, The Pennsylvania State University, State College, PA, USA
    Search for more papers by this author

Joy L. Pate, 324 Henning Building, University Park, PA 16802, USA.
E-mail: jlp36@psu.edu

Abstract

Citation Pate JL, Toyokawa K, Walusimbi S, Brzezicka E. The interface of the immune and reproductive systems in the ovary: lessons learned from the corpus luteum of domestic animal models. Am J Reprod Immunol 2010

The dynamic changes that characterize the female reproductive system are regulated by hormones. However, local cell-to-cell interactions may mediate responsiveness of tissues to hormonal signals. The corpus luteum (CL) is an excellent model for understanding how immune cells are recruited into tissues and the role played by those cells in regulating tissue homeostasis or demise. Leukocytes are recruited into the CL throughout its lifespan, and leukocyte-derived cytokines have been found in corpora lutea of all species examined. The proinflammatory cytokines inhibit gonadotropin-stimulated steroidogenesis, profoundly stimulate prostaglandin synthesis by luteal cells, and promote apoptosis. However, there is mounting evidence that leukocytes and luteal cells communicate in different ways to maintain homeostasis within the functional CL. Domestic animals have provided important information regarding the presence and role of immune cells in the CL.

Introduction

The discipline of reproductive immunology emerged from studies of the phenomenon by which the fetal allograft is tolerated by the maternal immune system, and it brought together researchers trained in basic reproductive physiology with those trained as immunologists in the more classical sense. Surely, the paradox of the fetal allograft remains at the forefront of reproductive immunology research and while great discoveries have been made, the mechanisms by which the fetus escapes rejection by the mother continue to engage the curiosity of reproductive immunologists. Meanwhile, a much broader field of reproductive immunology has emerged and includes studies of physiological processes in all reproductive tissues, as well as pathological conditions that impact fertility, autoimmunity, sexually transmitted diseases and reproductive cancers. The function of reproductive tissues may be modified by circulating immune cells and the cytokines they produce, but a feature common to most reproductive tissues is the presence of resident immune cells that may also be integrally involved in the regulation of tissue functions, including those of the gonads.

The corpus luteum (CL) is an ephemeral tissue, and its hormonal regulation is well characterized. The dynamic nature of luteal development, function and regression, coupled with the ability to precisely manipulate these processes, makes the CL an excellent model for understanding how immune cells are recruited into tissues and the role played by those cells in regulating tissue homeostasis or demise. The CL of domestic ruminants shares common features with the human CL. Both are derived from ovulation of a single, dominant follicle and develop into a fully functional CL in every cycle. The lifespan of the CL is very similar in humans and ruminants, allowing for ready comparisons to be made along the timeline of development, maintenance and regression. In contrast, rodents ovulate multiple follicles and the CL that are formed do not reach a plateau of function, but regress almost immediately after differentiation. Studies of the CL using rodent models employ hormones or physical manipulation to induce pseudopregnancy. However, the corpora lutea of pseudopregnant rodents secrete estrogen, similar to the human CL, whereas little to no estrogen is secreted from ruminant CL. This emphasizes the importance of choosing an appropriate animal model to answer a particular question. The differences that exist in luteal function among all the species also underscore the need to obtain basic information about gonadal function using diverse animal models. This review will be presented in the context of how domestic animals have been used to provide insight into the presence and role of leukocytes in luteal function and will review the more recent literature regarding immune cells in the ovary. For more comprehensive reviews, the reader is referred to previous reviews by the lead author1–4 as well as those of other well-known scholars in the field.5–7

Leukocytes in follicular development, ovulation and luteinization

Bukovsky and Presl8 proposed a role for the immune system in ovarian function and in 1980, Espey9 hypothesized that ovulation of ovarian follicles involves an inflammatory reaction. Evidence has since accumulated that supports these hypotheses, as well as indicating a potential role for immune cells and/or cytokines in the processes of follicular development, ovulation and luteinization. Cyclic variation in follicular chemokine expression implies that the ovary directs the recruitment of leukocytes to the follicle.10 Microarray analyses of bovine follicles indicated that immune response genes, including genes for chemokines and cytokines, were differentially expressed in follicles of different sizes, suggesting a role for these molecules in follicular development.11,12 Rats treated with granulocyte macrophage colony-stimulating factor (CSF2) had a greater number of large, pre-antral follicles compared to control rats.13 During the pre-ovulatory phase in women, monocyte chemoattractant protein (CCL2) is upregulated in the stroma surrounding the follicle, and macrophages are recruited during the ovulatory process.14 T lymphocytes are also recruited to the pre-ovulatory follicle. An interesting observation made using mice was that rare CD8αα+ T cells migrate into the pre-ovulatory follicle in response to the chemokine, thymus-expressed chemokine (TECK), and those cells are necessary for ovulation.15 Interleukin (IL) 8 induces neutrophil accumulation during ovulation in the rabbit16 and in the rat, human chorionic gonadotropin (hCG) induces an increase in the chemokine, fractalkine (CX3CL1), which enhances gonadotropin-stimulated progesterone production in cultured follicles and granulosal cells.17

Because of the ready availability of granulosal cells and follicular fluid from follicles of women undergoing IVF procedures, the presence of immune cells and cytokines in follicles can be measured and correlated with pregnancy outcome in humans. Macrophage colony-stimulating factor (CSF1) and its receptor are expressed in, and increase steroidogenesis by, human granulosal cells.18,19 Granulosal cells also produce the chemokine CXCL12 and express its receptor, CXCR4.20 These authors demonstrated that CXCL12 could induce migration of CD3+ T cells, but not B cells or natural killer cells, and suggested that recruitment of T cells to the follicle was associated with decreased apoptosis of granulosal cells. Women with endometriosis have reduced pregnancy rates following IVF compared to those with tubal infertility. The reduced fertility may be partially owing to the immune response mechanisms in the follicle, because concentrations of regulated upon activation, normal T cell expressed and secreted (RANTES) and CCL2 were altered in follicular fluid of women with endometriosis.21 Cytotoxic peptides, such as defensins, have been detected in human follicular fluid, but there was no correlation between defensin concentration and fertility in those patients.22 Pregnancy outcomes after IVF were greater in women who had fewer NKT (CD56+ CD3+) cells and more immunoregulatory NK (CD56+ CD16) cells.23 In summary, data collected using multiple species – rodents, rabbits, cows and humans – collectively point to an important role for leukocytes in follicular development and ovulation; however, domestic animal models have been less well studied in regard to the involvement of the immune system in follicular development.

Inflammation includes increased capillary permeability, leading to accumulation of fluid. It could be that the inflammatory process that characterizes ovulation results in leaking of leukocytes, such as macrophages, eosinophils and neutrophils into the surrounding tissue. This could be a non-specific accumulation that is rapidly cleared, or they could have a role in resolving the inflammatory process, which would be critical to ensure proper development of the CL. Cellular FLICE-like inhibitory protein (cFLIP), an antiapoptotic protein, is present in macrophages, steroidogenic and endothelial cells in greatest concentrations in the developing CL and may confer protective effects against apoptosis to ensure survival of the tissue despite high numbers of macrophages and granulocytes in the tissue.24 In CSF2-deficient mice, there is increased activation of ovarian macrophages, which coincides with compromised luteinization and steroidogenesis in early pregnancy.25 There is also evidence that leukocytes are involved in the process of luteinization. It has also been suggested that eosinophils influence the process of angiogenesis during luteal development.26,27 The chemokine responsible for the selective migration of eosinophils may be Substance P, because it is produced in the developing CL, along with its receptor, neurokinin-1 receptor.28 Although eotaxin was proposed to play this role, Vogel et al. (2005)29 detected only very minimal amounts of eotaxin mRNA in the bovine ovary.

Identification of immune cell types in the CL

Immune cells are present in the CL of all species studied thus far. Many studies have focused on the increase in immune cells observed during luteolysis, but immune cells are present in the CL throughout its lifespan. Immunohistochemical analyses reve-aled that the type of leukocytes changes with functional state of the CL. In the bovine CL, there appears to be an inverse relationship in the number of granulocytes and lymphocytes throughout the lifespan of the CL. Eosinophils are more abundant in developing than fully functional CL30,31 and appear to be recruited into the developing CL by expression of P-selectin (CD62P) on endothelial cells.32 In contrast, T lymphocytes are undetectable or exist in very low number in the developing CL, but the number of T cells is greater in a fully functional CL and further increases during luteolysis.33,34 T lymphocytes are generally categorized into the major classes of CD4+, CD8+ or γδ+ cells. Temporal changes in each of these classes of T cells might lead to suggestions about their roles in luteal function. However, use of immunohistochemical techniques to quantify these cells in bovine luteal tissue sections has led to discrepant results. Bauer et al. (2001)33 and Townson et al. (2002)34 observed greater numbers of both CD4+ and CD8+ cells in regressing CL, suggesting a role for both cell types in luteolysis. However, Penny et al. (1999)30 observed no change in the number of CD4+ T cells. Further, Bauer et al. (2001)33 reported that CD8+ cells were found in proportionally greater number than CD4+ cells, whereas Townson et al. (2002)34 found no significant difference in the proportion of CD4+ and CD8+ cells. Davis and Pate (2007)35 reported the presence of γδ+ T cells in midcycle and regressing CL, but quantitative analysis was not performed in that study. In an effort to better understand changes in T-lymphocyte populations in the bovine CL, we recently undertook a study to isolate resident T cells from the CL and characterize them via flow cytometry. More CD8+ than CD4+ T cells were observed in the luteal-resident T-cell population using this method. It was also apparent that specific subpopulations of T cells are selected for migration into the CL, and changes in subsets of the major classes of T cells may be more reflective of their functional significance than overall proportions of CD4+, CD8+ or γδ+ cells.36 Macrophages also increase in the bovine CL during luteal regression.33–35

Although less well studied than the cow, the horse, sheep and pig have also been used to characterize changes in immune cells during the lifespan of the CL. Similar to the cow, T cells were greater in the equine CL late in the estrous cycle and administration of prostaglandin (PG)F resulted in an increase in CD8+ cells.37 Studies using the sheep and pig have shown that eosinophils were greater in the developing CL, but eosinophils38–40 as well as macrophages41 migrate into the CL during luteolysis, and two macrophage phenotypes may be distinguished based on presence or absence of cytoplasmic lipid droplets.42 Neutrophils are present in much lower number than eosinophils. T lymphocytes also increase during luteolysis in mice,43,44 but no changes were observed during pregnancy or pseudopregnancy in rats.45

Rabbits have served as an excellent model to study changes in T cells and macrophages related to luteal function. T cells are found within the CL prior to the recruitment of macrophages,46 but it is the macrophages that increase in number during luteal regression47,48 or following estradiol withdrawal.49 In the rabbit, it seems that T cells infiltrate into the CL first, followed by macrophages.47 However, in the mouse, macrophages decline in number in the latter stages of luteal regression, whereas T lymphocytes continue to increase.43 Macrophages derived from mouse ovaries are less active than circulating macrophages in the uptake of foreign molecules, suggesting that the tissue environment may regulate the function of resident macrophages.50 The recruitment of monocytes/macrophages into the CL of the rat is independent of the concentration of progesterone.50

Although Brannstrom et al.45 did not observe a change in the number of T cells in the human CL during the menstrual cycle, most reports characterizing lymphocytes in human CL indicated that T cells increase in the late luteal phase,51–53 similar to what has been reported for the bovine CL. Also similar to bovine CL, B cells and NK cells are undetectable or in very low number in human CL,51–54 whereas eosinophils and neutrophils are greater in developing CL.27,52,53 Macrophages are present in the human CL, but it is not clear whether they exist in greater number in the early CL,55 or continue to increase to a maximum in the late luteal phase.56 Of note is that fewer macrophages are present in the CL after in vivo treatment with hCG to simulate early pregnancy.56

Recruitment of leukocytes into the CL

The dynamic nature of leukocyte populations in the CL suggests that production of chemokines or temporal changes in endothelial cell adhesion molecules must regulate the transmigration of leukocytes into the luteal parenchyma. Surprisingly, a limited amount of work has been performed in this area. In the cow, most of the work on chemokines has focused on CCL2, which appears to play a pivotal role in the recruitment of macrophages and lymphocytes into the CL. The mRNA for CCL2 is greater in regressing compared to functional CL57,58 and CCL2 protein increased with age of the CL; concentration of CCL2 and its mRNA were correlated with accumulation of monocytes, macrophages and T lymphocytes.34,57 The expression of CCL2 on luteal endothelial cells was stimulated by tumor necrosis factor alpha (TNF) and interferon gamma (IFNG).59 Interestingly, CCL2 was upregulated in day 11, but not day 4, CL by injection of PGF, implying a potential role of this chemokine in the acquisition of luteolytic capacity.60

Human granulosal-lutein cells secrete IL8, which could serve as the chemoattractant for the neutrophils that infiltrate the developing CL.61 Although the horse is not as commonly used for studies of ovarian function, Lawler et al. (2002)62 used culture medium collected after incubation of luteal tissue to demonstrate that the equine CL produces chemokines for both mononuclear and polymorphonuclear cells in late diestrus and during luteolysis, but not during the midluteal phase.

Cytokine expression and actions in the CL

Tumor necrosis factor-α is produced in the CL and may have a functional role in luteolysis. TNF concentrations were greater in regressing CL and corresponded to the accumulation of macrophages.46,47 Luteal concentration of IL10, which is considered to be an antiinflammatory cytokine, increases at the same time as TNF in rabbit CL,48 leading those authors to speculate that antiinflammatory cytokines are necessary to control immune cell function during luteolysis, a concept that was also proposed by Pate and Keyes (2001).4 Around the same time, work began on expression and function of cytokines in the bovine CL. The mRNA for TNF30,63,64 and IFNG63 are detectable in bovine CL. Although the luteal concentration of TNF mRNA does not vary significantly during the estrous cycle, the protein is greater on estrous cycle days 13–18.64 The ability of TNF to elicit a physiological response in the CL is suggested by the presence of TNF receptor type I (TNFRI). Interestingly, fewer TNFRI receptors were present in the day 15–17 CL, which is the time that corresponds to the greater concentrations of TNF, suggesting that TNF may downregulate its own receptor in this tissue.64 Shaw and Britt65 measured the secretion of luteal TNF using continuous flow in vivo microdialysis. The cytokine was only detectable in the dialysate following the decline in progesterone, indicating that TNF may be secreted from cells into the intercellular spaces to facilitate luteal regression. The first functional effects of TNF on luteal cells were demonstrated using bovine cells, in which TNF increased production of prostaglandins, decreased LH-stimulated progesterone synthesis and upregulated class I, but not class II, major histocompatibility (MHC) molecules,66 as well as a proteosomal subunit responsible for processing of MHC class I-presented peptides, LMP10.67 Although these effects of TNF would be consistent with a role in luteolysis, there is some evidence that higher concentrations of TNF promote luteal survival.68,69

Using cultured bovine luteal cells, it was demonstrated that IFNG has effects similar to those of TNF. Interferonγ inhibits LH-stimulated progesterone production, increases prostaglandin synthesis, upregulates class I and class II MHC molecules and causes cell death.70,71 The apoptotic effects of IFNG on luteal cells involve upregulation of Fas expression72 and are mediated by the action of indoleamine 2,3-dioxygenase, which essentially starves the cells of tryptophan.73 The effects of IFNG on luteal cells are potentiated in the presence of TNF,66,72 probably because of TNF-induced upregulation of IFNG receptors.74 Induction of apoptosis by IFNG + TNF can be suppressed by IFNA or acetylcholine,75,76 and progesterone may confer some protective effect against apoptosis in the CL.77

Additional cytokines that have been detected in the bovine CL include interleukin 1α (IL1A), IL1β (IL1B), transforming growth factor (TGFB1), macrophage migration inhibitory factor (MIF) and osteopontin (SPP1). Both interleukins stimulate prostaglandin production by luteal cells, but have no effect on progesterone synthesis.78,79 Interleukin 6 decreased the progesterone production by porcine luteal cells, and IL6 receptor mRNA was increased in the regressing compared to the functional CL.64 The greatest amount of MIF mRNA was found in the developing CL and MIF was localized to large luteal cells, suggesting a potential role in differentiation of granulosal-luteal cells in addition to its role as a proinflammatory cytokine.80 Another proinflammatory cytokine, SPP1, is present in the bovine CL81 and may play a role in paracrine communication between luteal cells and T cells.82 Bovine luteal cells in vitro secrete large amounts of TGFB1.83 Treatment of luteal cells with TGFB1 reduces progesterone secretion, and the actions of TGFB1 are likely mediated by the activation of the early growth response (EGR)1 protein, which is upregulated during PGF-induced luteolysis.84 Overall, since the late 1980s, the bovine CL has served as a useful model to study cytokines in luteal function, and the data collectively point to a role of the proinflammatory cytokines in luteolysis. However, more work needs to be undertaken to determine functional interactions of the complex array of cytokines found within the CL.

A number of investigators have obtained similar results using rodent and primate CL. Interleukin 1A, IL1B and IL18 were detected in the murine CL,85,86 and IL1B stimulated the prostaglandin production in luteal cells from rats.87 The reports of effects of IL1B on progesterone production have been inconsistent. Interleukin 1B has been reported to suppress progesterone synthesis in CL of rats,87 stimulate progesterone synthesis in human luteal cells88,89 or have no effect in bovine luteal cells.78,79 While it is tempting to speculate that there are species differences in the effects of IL1B on steroidogenesis, it is likely that the differences are a result of the in vitro conditions. In fact, using human luteal cells, Castro et al.90 and Kohen et al.91 demonstrated that completely opposite effects of IL1B on progesterone production could be obtained depending on the presence or absence of luteal leukocytes in the cultures. In those studies, leukocytes appeared to mediate the antisteroidogenic effects of IL1B, because IL1B stimulated steroidogenesis in leukocyte-depleted cultures. In rodents and primates, the expression and functional effects of TNF and IFNG appear to be similar to what has been observed in the cow. No profound changes were noted in concentrations of TNF throughout the lifespan of the murine CL, while IFNG was detected only in regressed CL.43 Similar to what had been observed in bovine luteal cells,66 TNF inhibited progesterone production in luteal cells from rats, and this action was a result of downregulation of steroidogenic acute regulatory protein (StAR) and luteinizing hormone (LH) receptor.43 Also similar to the cow, the combination of TNFA and IFNG exert profound proapoptotic effects on rodent luteal cells92,93 and alter the Fas/FasL system.43,94 In macaque CL, TNF mRNA increases with age of the CL,95 and human luteal cells collected during the late luteal phase secrete greater concentrations of TNF in vitro than cells from the early luteal phase.96 Further, luteolytic effects of TNF and IFNG on human and non-human primate luteal cells also appear to involve Fas/FasL. Using mutant mouse models, Henkes et al.97 provided strong evidence for a role of TNF, mediated by increased activity of acid sphingomyelinase, in luteolysis. Thus far, similar types of single-gene mutant domestic animal models have not been developed, providing one limitation in studying pathways that mediate cytokine action in these species.

Role of immune cells in luteolysis

The expression and functional effects of the proinflammatory cytokines described above point to a role in luteolysis. The mRNA of the three most well-studied cytokines in regard to luteal function, IL1B, TNF and IFNG, are upregulated during luteolysis in the cow,98 as is TNF receptor I.99 It is likely that multiple cell types are targets for the actions of TNF in the regressing CL. As previously stated, TNF, either alone or in combination with IFNG, inhibits LH-stimulated progesterone production while stimulating prostaglandin production by steroidogenic cells. Conversely, the apoptotic effects of TNF may be exerted primarily on luteal endothelial cells because TNF-induced apoptosis in cultures of luteal endothelial, but not steroidogenic, cells.99 TNF-induced upregulation of chemokine secretion by endothelial cells100 may be an early signal to recruit leukocytes into the CL to facilitate luteolysis. In fact, despite being the known luteolytic agent in ruminants, PGF does not upregulate CCL2 secretion by cultured luteal endothelial cells nor does it mediate the onset of apoptosis in luteal steroidogenic or endothelial cells.3,101 However, chemokine secretion by luteal endothelial cells is stimulated by coculture with peripheral blood leukocytes (PBLs), which are presumably doing so via secretion of cytokines such as TNF.10,101 These authors clearly demonstrate that complex interrelationships among immune cells, endothelial cells and steroidogenic cells are necessary for PGF-induced luteal regression.

In all species studied, there are greater numbers of macrophages and lymphocytes in the regressing CL than in the functional CL. This is primarily because of the transmigration of leukocytes from the capillaries into the parenchymal tissue. However, the percentage of proliferating leukocytes increases from 20% in a fully functional CL to 70% during luteal regression.33 Most of the proliferating cells are macrophages, although proliferation of a small proportion of lymphocytes was also observed in that study. This raises the very interesting question of not only what factors recruit leukocytes into the CL, but what causes the resident leukocytes to be activated and proliferate when the CL undergoes regression.

It has been suggested that one activator of luteal monocytes is prokineticin 1, which is found in greater abundance in the regressing than functional CL.102 T lymphocytes may also be activated by altered expression of, or peptide presentation by, MHC molecules. Class I MHC molecules are constitutively expressed on all luteal cells, whereas expression of class II MHC molecules is restricted to subpopulations of luteal cells and is increased late in the estrous cycle and during luteolysis.67 However, expression of MHC molecules, costimulatory molecules, and the intracellular components of the peptide processing machinery for presentation by MHC are all present long before luteolysis and T-cell activation occurs.67,103,104 In fact, the mRNA for the CD80 and CD86 costimulatory molecules and DMα, which is a component of the MHC class II antigen processing machinery, are found in greater concentrations in midcycle compared to late cycle CL.103,104 Interestingly, the IFNG-inducible proteosomal subunit, LMP10, is upregulated in luteal cells by TNF, whereas DMα is downregulated.103 Also, class II MHC expression on luteal cells is reduced in vivo during maternal recognition of pregnancy105 and in vitro by combined treatment with progesterone and IFNA.2

Immune cells and homeostasis within the CL

Bovine luteal cells are potent stimulators of T-cell proliferation in vitro.106 Luteal cells collected from regressing CL, which possess a greater number of class II MHC molecules are more potent stimulators of T-cell proliferation than midcycle luteal cells, and luteal cell-induced T-cell proliferation is inhibited by high concentrations of progesterone.67,107 Blocking costimulatory molecule function with antibodies results in a reduction in luteal cell-induced T-cell proliferation.104 Collectively, all these results point to a role for MHC molecules in the activation of T cells within the CL during luteolysis. However, a number of more recent observations have led to an expanded hypothesis about the interactions between luteal cells and lymphocytes in the CL. First, the addition of anti-MHC blocking antibodies to luteal cell–T cell cocultures only partially suppresses T-cell proliferation.35,106 Secondly, the majority of the MHC molecules were localized to subsets of capillary endothelial cells in situ, with similar expression in CL collected on day 10 and day 18 of the estrous cycle.104 Cytokeratin-negative bovine luteal endothelial cells express class II MHC molecules in culture,100 whereas cultured luteal steroidogenic cells express only minimal class II MHC in the absence of IFNG.70 Finally, we discovered that the T cells that are most responsive to luteal cells are γδ+ T cells, which are not MHC restricted.35 The responses of γδ T cells tend to be driven more by imbalance within the host tissue, requiring γδ T-cell autoreactivities to be tightly controlled.107 Most information on tissue-specific γδ T cells has been obtained from studies of these cells localized within skin and gut epithelium (known as intraepithelial lymphocytes, or IELs), but a distinct subset of γδ T cells comprise about one-half of all T cells within reproductive epithelia108,109 and are thought to be important for non-rejection of the fetal allograft. In fact, γδ T cells within the reproductive tract are increased nearly 100-fold in pregnant animals compared to non-pregnant animals,110 and more γδ T cells are found in the peripheral blood of healthy pregnant women compared to that of non-pregnant women or recurrent aborters.111

Although the stimulation of T cells by luteal cells may lead to activation of a proinflammatory response during luteolysis, it is equally plausible that luteal cells and resident immune cells interact in such a way as to regulate homeostasis within the tissue during periods of tissue remodeling and/or extremely high metabolic activity. The interaction of multiple cell types, including immune cells, to maintain homeostasis within the ovary was proposed by Bukovsky et al.112 and was termed the ‘tissue control system’. A role for immune cells to abate an inflammatory response in the ovary was postulated by Pate and Keyes.4

Conclusion

In summary, leukocytes are recruited into the CL throughout its lifespan, but the specific types of leukocytes found within luteal tissue vary depending on the functional state of the CL. Cytokines that are produced primarily by lymphocytes and macrophages have been found in corpora lutea of all species examined and have profound effects on luteal cells in vitro. Interactions between leukocytes and luteal parenchymal cells (steroidogenic and endothelial cells) result in functional changes in both the parenchymal cells and the leukocytes, suggesting that bidirectional communication is an important component in overall regulation of tissue function and is likely as important in homeostasis as in regression of the CL. A summary of the proposed interactions between ovarian parenchymal cells and leukocytes is depicted in Fig. 1. The ephemeral nature of the CL and the ability to hormonally control its development and demise make it a relevant tissue in which to study dynamic changes in tissue resident leukocytes. Although it is critical that diverse species be studied to develop a comprehensive picture of the role of leukocytes in luteal function, the size and accessibility of the bovine CL provides an exceptionally useful model for these studies.

Figure 1.

 (a) Leukocytes and their products participate in the process of ovulation. LH-induced biochemical changes within the thecal layer initiate a series of changes in the basement membrane as well as leakage in the thecal vasculature. Chemokines (CCL2, CXCL12, TECK and CSF2) present in follicular fluid of pre-ovulatory follicles coupled with increased expression of adhesion molecules on endothelial cells facilitate recruitment of specific leukocyte subpopulations. Leukocytes secrete products that facilitate rupture of the follicular wall, tissue remodeling and proliferation of luteal cells after ovulation. (b) Interaction between leukocytes and luteal cells in a functional and regressing corpus luteum (CL). Infiltrating leukocytes in the CL may assume different roles relative to the state of the CL. The activated immune cells expressing ligands to endothelial adhesion molecules are recruited into the CL in response to chemokine mediators IL8 and CCL2, which are produced by endothelial cells. In a functional CL, immune cells may potentiate survival by producing antiapoptotic molecules like cFLIP. During regression, in response to PGF, increased expression of inflammatory cytokines (TNF, IFNG) inhibit P4 production and activate endothelial cells to recruit more immune cells. IL1B, IFNG and TNF may also activate resident T cells with upregulation of FasL. FasL induces apoptosis in Fas-expressing luteal cells. Inflammatory cytokines also induce apoptosis in both luteal cells and endothelial cells. Bidirectional communication between luteal and immune cells may thus facilitate hormonally regulated survival and demise of the CL. Solid lines represent bidirectional communication whereas dashed lines depict secreted molecules. LLC, large luteal cells; SLC, small luteal cells; LEC, luteal endothelial cells; Mac, macrophage/monocyte; IFNG, interferon gamma; TNF, tumor necrosis factor.

Acknowledgments

The authors thank Ms. Melanie Boretsky for her assistance with the preparation of this manuscript.

Ancillary