Immunomodulatory effects of sex hormones: requirements for pregnancy and relevance in melanoma

Abstract

Similarities between the pathologic progression of cancer and the physiologic process of placentation (e.g., proliferation, invasion, and local/systemic tolerance) have been recognized for many years. Sex hormones such as human chorionic gonadotropin, estrogens, progesterone, and others contribute to induction of immunologic tolerance at the beginning of gestation. Sex hormones have been shown to play contributory roles in the growth of cancers such as breast, prostrate, endometrial and ovarian cancer, but their involvement as putative mediators of the immunologic escape of cancer are still being elucidated. Herein, we compare the emerging mechanism by which sex hormones modulate systemic immunity in pregnancy and their potentially similar role in cancer. To do this, we conducted a PubMed search using combinations of the following key words: immune regulation, sex hormones, pregnancy, melanoma and cancer. We did not limit our search to specific publication dates. Mimicking the maternal immune response to pregnancy, especially in late gestation, might aid in design of better therapies to reconstitute endogenous anti-tumor immunity and improve survival.

INTRODUCTION

In 1948, Beard and Krebs acknowledged a striking similarity between a trophoblast and a tumor, publishing their observation titled “The Unitarian or Trophoblastic Thesis of Cancer.”^1,2 Since then, these similarities have been extensively studied; many shared pathways and immunologic mediators have been identified.^3,4 The purpose of this review is to take an in-depth look at existing research describing the role of sex hormones in the potentially parallel settings of reproductive and tumor immunology, with a focus on metastatic melanoma. Though imperfectly understood, sex hormones are important regulators of the immune system in both pregnancy and cancer.^5,6 It is clear that they are involved in regulation and modifications of the immune system to allow invasion, proliferation and migration of tumor cells and trophoblasts.⁴ It is possible that an organ system-level view of the process of placentation as well as melanoma progression could yield additional insights into potential therapeutic targets for hormone-based immune modulation.⁷

The complexities and redundancies involved in orchestration of the maternal response to pregnancy as well as the host response to cancer are increasingly appreciated.^3,8 Importantly, however, we and others have observed that neither pregnancy nor advanced cancers are static immunologic events.^9–11 Oscillations in systemic immunity between inflammation and tolerance seen in patients with metastatic melanoma have been documented and seem to follow a biologically predictable pattern. When tolerance seen in malignant melanoma is disrupted and brought back to a state of inflammation, patients have a much better prognosis than those whose immune systems stay in an immunologically exhausted state. Pregnancy is also characterized by many hormonal fluctuations, although the time scale is for these hormonal and immunologic changes may be measured in weeks as opposed to days.¹² (figure 1) While it seems intuitive to consider the involvement of sex hormones interacting with the maternal immune system during pregnancy, it is less obvious but just as possible that such hormones alter systemic immunity in the setting of cancers such as melanoma. We describe the observations and experimental evidence supporting such involvement in the following paragraphs.

Figure 1.

Hormonal changes in week of hormones important for the regulation of gestation in healthy pregnant women. CRH = corticotropin releasing hormone; hCG = human chorionic gonadotropin.

CLINICAL EVIDENCE OF HORMONAL REGULATION OF MELANOMA

The skin is capable of producing many neuroendocrine mediators such as melanin, steroids, thyroid hormones and sex hormones such as androgen, estrogen and progestin in order to maintain homeostasis; any failure to communicate between the skin, endocrine and immune system could result in deregulation and disease.^13,14 Both melanocytes and melanoma tumors produce pigment in the melanosome that protects the skin against damaging ultraviolet rays through positive regulation by hormone like L-tyrosine and L-dihydroxyphenylalanine.^15,16 Although the interplay between sex hormones and the immune system in melanoma remains poorly understood, several clinical observations support the role of sex hormones in melanoma development. Melanomas that are responsive to estrogens are associated with the superficial spreading melanoma subtype, a type of tumor with a much better prognosis. Additionally, estrogen exerts a proliferative effect on melanocytes and can lead to the development of hyperpigmentation in women using oral contraceptives (OC) or hormonal replacement therapy (HRT).¹⁷ Whether hormonal contraceptives increase the risk of melanoma is a matter of ongoing debate. Koomen et al has reported that high estrogens increase a woman’s risk for developing malignant melanoma,¹⁸ while Lens and Bataille have not observed a significant association.¹⁹ It may be no coincidence that melanoma is the most common form of cancer associated with pregnancy.²⁰ This is believed to be due to the trophoblast’s increased need for lymphangiogenesis, which the melanoma then uses to promote its own growth. Complementary to this hypothesis, demographics and incidence may also provide an explanation as to this phenomenon. Additionally, pregnant women are more likely to be diagnosed with an invasive melanoma than their non-pregnant counterparts.

We have shown that aging in healthy subjects is associated with a Th2 bias.²¹ Women are more likely than men to develop melanoma before age 40, after which diagnosis of melanoma is observed at a much higher rate in males.²² However, females diagnosed with melanoma have a better prognosis than males, and premenopausal women have higher survival rates compared to postmenopausal women.^23,24 Interestingly, melanoma metastasizes at a much slower rate in women compared to men, and the pattern of metastatic spread is also different, with more loco-regional recurrences observed in women.^25,26 Still, reasons behind these sex differences have yet to be elucidated. With an extensive amount of emerging data demonstrating the importance of sex hormones on immune function, in the remainder of the review we will explore the mechanisms behind endocrine sex hormone mediated immunomodulation. First, we will highlight the potential involvement of hormones in tumor promotion, both via tolerance induction as well as chronic inflammation and angiogenesis. Next, we will describe hormones that may help improve anti-tumor immunity. Finally, we briefly discuss sexual dimorphism in immune responses, which may have implications for the development of personalized immunotherapy. A better understanding of immunological switches that control tolerance, immune activation, and immune reconstitution, of which all can be studied using the different phases of pregnancy as a model, could result in novel immunologic treatment strategies for melanoma and other malignancies.

PUTATIVE PRO-TUMOR/IMMUNE SUPPRESSIVE HORMONES: HUMAN CHRONIC GONADOTROPIN, PROGESTERONE, PLACENTAL GROWTH FACTOR, AND RELAXIN

Human chronic gonadotropin

Soon after fertilization, the embryonic blastocyst begins secreting human chronic gonadotropin (hCG). HCG is a glycoprotein hormone primarily produced by trophoblasts, which promotes many processes including implantation, recognition, differentiation, angiogenesis and fetal-maternal homeostasis.²⁷ (figure 2) Levels of hCG continue to rise until the 11^th week of gestation and then slowly decreases through the remainder of the pregnancy. The main purpose of hCG in pregnancy is to prevent degradation of the corpus luteum and stimulate progesterone production.²⁸ HCG also plays an important role in promoting immune suppression in the decidua by preventing maternal macrophage phagocytosis to the invading trophoblast in order to establish immune tolerance.²⁹ The early trophoblast promotes fetal tolerance by secreting hCG which acts as a powerful chemoattractant for T regulatory cells (Treg) to migrate to the placenta after fertilization.³⁰ Treg cells play an important role in maintaining self-tolerance and modulating tolerance to non-self antigen, such as those displayed by the fetus. Expression of complement component 3 (C3) was found to be upregulated by hCG on stromal cells of the baboon endometrium post ovulation, suggesting hCG is also able to modulate the decidual environment during the preimplantation stage.^31,32

Figure 2.

Effects of sex hormones during early pregnancy to promote implantation and fetal tolerance and the proposed role in initiation of cancer. CRH = corticotropin releasing hormone; E2 = estradiol; hCG = human chorionic gonadotropin; IFN = interferon; IL = interleukin; uNK = uterine natural killer cell; VEGFR = vascular endothelial growth factor receptor.

HCG also regulates uterine natural killer (uNK) cells.³³ This uNK subset makes up approximately 70% of the lymphocyte population in the endometrium and plays an important role in maintaining and regulating the uterine spiral arteries during the first trimester.³⁴ UNKs stimulate decidual monocytes to secrete IFN-γ promoting Treg proliferation through indoleamine 2,3-Dioxygenase (IDO) and transforming growth factor beta (TGF-β).³⁵ Interestingly, we have found uNK cells in the blood of patients with stage IV melanoma and they also positively correlated with TGF-β levels in plasma.³⁶ HCG drives a systemic response during pregnancy. Before in vitro fertilization (IVF), women given hCG had increased levels of anti-inflammatory IL-27 and IL-10, reduced levels of pro-inflammatory IL-17, which resulted in elevated Tregs and a more receptive uterine wall for implantation.³⁷ They also notice that hCG affected the maternal adaptive immune system, promoting a Th2 differentiated state by activating T cells that produce IL-4 while inhibiting T cells that secrete IFN-γ.³⁷

Additionally, the formation of new blood vessels is also driven by hCG which acts upon pro-angiogenic molecules such as vascular endothelial growth factor receptor (VEGFR)-1 and angiopoietins.³⁸ Soluble VEGF-C aids in immune tolerance by suppressing the cytotoxic activity of uNK cells at the fetal-maternal interface.³⁹ HCG enhances VEGF production by endometrial cells through paracrine feedback, further promoting blood vessel formation to the fetus.⁴⁰ IL-10, an inhibitor of inflammatory cytokines which is found at high levels near the beginning of pregnancy, induces the trophoblast to produce VEGF-C and stimulate placental angiogenesis.⁴¹ Also, endocrine gland-derived vascular endothelial growth factor (ED-VEGF) has been implicated as a negative regulator of trophoblast invasion in the placenta, as high levels of ED-VEGF are seen in preeclampsia, and has recently been hypothesized to be regulated by hCG.⁴² HCG prevents apoptosis of endometrial cells though the Fas-FasL pathway, which in turn drives maternal tolerance during early pregnancy.⁴³

HCG plays very similar roles in the invasion and progression of cancer (melanoma). The hCG receptor was first discovered in trophoblastic neoplasms, which suggested that it has an important role in regulating growth and invasion in not only in pregnancy, but in cancer as well.⁴⁴ Several studies have found that many cancers, including bladder cancer, cervical cancer, lung cancer, pancreatic cancer and colorectal cancer can be diagnosed by high levels of hCG in the serum.^45,46 HCG mRNA has also been expressed in tumor cells from patients diagnosed with malignant melanoma and has been suggested for use as a biomarker for disease.⁴⁷ Others have shown that up to 60% of active neoplasia will express high levels of hCG in serum.⁴⁸ Antibodies against hCG have been detected during malignancy, but hCG levels are so high that the effect of these antibodies seems to be minimal, therefore allowing hCG to act as an autocrine growth factor to promote malignancy.⁴⁹ Tumor secretion of hCG prevents apoptosis, allowing the cancer to become more resistant and aggressive.⁵⁰ As a potent angiogenic factor, hCG secreted by the tumor stimulates sprout formation through vasodilation, maturation and increased vessel permeability, thus promoting tumor growth.⁵¹

Progesterone

Initial progesterone production in pregnancy is induced by hCG. Progesterone secretion by the placenta continues to increase throughout pregnancy, decreasing approximately 4 weeks before labor onset. Progesterone is a potent immunomodulator that establishes Th2 bias in pregnancy by reducing the production of pro-inflammatory cytokines by macrophages in response to infectious stimuli, and altering cytokine secretion of T cell subsets towards IL-10 production.⁵² (figure 2 and 3) Further, it induces secretion of chemokines such as CXCL10, CX3CL1 and CCL2 that localize Th2 biased immune cells to the placenta and upregulates non-classical HLA-G.⁵³ Interestingly, most progesterone receptor (PR) expressing decidual T cells express a γ/δ T cell receptor (TCR).⁵⁴ This limits the number of ligands the TCR can recognize which provides protection to the growing fetus. These T cells are also thought to be able to identify antigens presented by trophoblasts.⁵⁵ The progesterone receptor (PR) comes in two main isoforms, PR-A and PR-B, which compete with one another for progesterone binding.⁵⁶ PR-B is expressed on myometrial cells during the majority of pregnancy and inhibits expression of pro-inflammatory genes; however, PR-A is expressed during labor and it in turn promotes pro-inflammatory gene activation.⁵⁷ PR-A overexpression at the end of pregnancy also promotes the activation of estrogen by increasing estrogen receptor alpha (ER-α) expression.⁵⁸ A smaller, soluble PR isoform, PR-C, has been found to compete specifically with PR-B for progesterone binding, by binding directly to progesterone and inhibiting PR-B signaling near parturition, therefore, promoting labor associated myometrium changes and activating pro-inflammatory pathways.⁵⁹

Figure 3.

Maintenance and regulation of tolerance by sex hormones during mid-pregnancy and the proposed role of hormones in progression/metastasis of cancer. CRH = corticotropin releasing hormone; C3 = complement component 3; ER = estrogen receptor; E2 = estradiol; E3 = estrone; hCG = human chorionic gonadotropin; IFN= interferon; IL = interleukin; PlBF = placental binding factor; TNF = tumor necrosis factor; VEGF = vascular endothelial growth factor.

The immune effects of progesterone are exerted by inhibition of pro-inflammatory transcription factor NF-κB through IκB.⁶⁰ Progesterone effects are also mediated by progesterone-induced blocking factor (PIBF), which is expressed on lymphocytes in the decidua and is important in maintaining pregnancy. Women who suffer from spontaneous miscarriages or have high stress levels have been found to have low PIBF levels, which was associated with pregnancy complications.⁶¹ Natural killer (NK) cells in pregnancy are very sensitive to progesterone: low levels of progesterone are required to achieve inhibition, compared to 100 fold higher levels needed to achieve similar NK cell inhibition in non-pregnant individuals.⁶² In mice, progesterone activates a Th2 immune biased response by inhibiting maturation of dendritic cells, which would more readily initiate a Th1 response through the promotion of cytotoxic T cell expansion.⁶³

In cancer, progesterone modulates immune responses by inhibiting T cell proliferation⁶⁴ and IFN-γ expression⁶⁵ while enhancing IL-4, IL-5, IL-6 and IL-10 production, and promoting a humoral response.⁶⁶ Interestingly, human melanocytes can produce steroids de novo through the metabolism of progesterone or cholesterol.^67–69 Progesterone has been shown to inhibit proliferation of human melanocytes by blocking estrogen effects.⁷⁰ The progesterone receptor has been identified in the cytoplasm and nucleus of melanocytes by immunohistochemistry.⁷¹ Melanoma cells that do not express the PR were still found to be regulated in the presence of progesterone, but are modulated through signal transduction versus transcription.⁷² Progesterone induced blocking factor (PIBF) has been suggested to play a role in cell cycle regulation and Th2 biased immunity through the IL-4 receptor.⁷³ PIBF mRNA is constitutively expressed in tumor cells and does not require the presence of the progesterone receptor, which provides a mechanism for cancer to escape anti-tumor immune responses.⁵⁵ In vitro experiments using WM266-4 cells showed that progesterone, 17beta-estradiol and dihydrotestosterone given together can inhibit tumor growth through the down-regulation of IL-8, which is a potent cytokine that induces melanoma cell growth.⁷⁴

Placental growth factor

We have previously described a VEGF-driven state of chronic inflammation in metastatic melanoma.⁷⁵ Placental growth factor (PlGF), a VEGF homologue, may play a similar role in perpetuating chronic inflammation. PlGF plays a significant role in embryogenesis, promoting both angiogenesis and vascularization to the fetus during inflammation.⁷⁶ It was discovered that Flt-1, a VEGF receptor, binds PlGF and mediates recruitment of monocytes.⁷⁷{ Peripheral blood monocytes (PBM) treated with PlGF showed increased expression of pro-inflammatory cytokines IL-1β and TNF-α, and chemokines MCP-1, IL-8 and MIP-1β, suggesting PlGF plays an important role in inducing inflammation.⁷⁸ Melanocytes and melanoma tumors are also known to secrete PlGF, which makes them weakly responsive to anti-VEGF therapy.⁷⁹ However, when PlGF was neutralized, even in tumors resistant to anti-VEGF therapy, the tumor could no longer signal through VEGFR-1, inhibiting growth during a pre-clinical study.⁸⁰ Therefore, it has been hypothesized that PlGF plays a significant role in allowing a tumor to become drug resistant.^81,82 Further studies need to be conducted to fully understand PlGF’s role in inflammation and tumor progression.

Relaxin

The role of relaxin in pregnancy (induction of MMPs, ECM remodeling, labor), as well as induction of pro-inflammatory cytokines IL-6 and IL-8, has been well documented in rhesus monkeys.⁸³ While relaxin has not been studied in melanoma to date, its role in carcinogenesis of breast and prostate cancer has been well established. Relaxin was implicated in tumor growth, cell invasion during pregnancy and likely tumor invasion in carcinogenesis as well as in angiogenesis by induction of VEGF.^84–86 Whether relaxin plays a role in melanoma development remains to be elucidated.

PUTATIVE ANTI-TUMOR/PRO-INFLAMMATORY HORMONES: CORTICOTROPIN-RELEASING HORMONE, PROLACTIN AND VISFATIN

CRH

Corticotropin-releasing hormone (CRH) has been postulated to regulate the duration of gestation, with its levels being the highest during labor. (figure 4) CRH works directly on myometrial cells to facilitate the onset of labor, and it is tightly regulated by progesterone, estrogens, nitric oxide, IL-1β and tumor necrosis factor alpha (TNF-α).⁵⁹ Corticotropin-releasing hormone binding protein (CRHBP) binds CRH and inactivates its ability to promote adrenocorticotropic hormone (ACTH) production.⁸⁷ CRHBP levels decrease throughout pregnancy allowing high levels of CRH to accumulate and promote labor. Other actions of CRH include regulation of fetal blood flow, placental prostaglandin and cortisol production and uterine contractility.⁸⁸

Figure 4.

Role of sex hormones in the promotion of inflammation and labor as a natural method for provoking regression of tumor in cancer patients. bFGF = basic fibroblast growth factor; CRH = corticotropin releasing hormone; CRHBP = corticotropin releasing hormone binding protein; E2 = estradiol; hCG = human chorionic gonadotropin; hPRL = human prolactin; IFN = interferon; IL= interleukin; PR = progesterone receptor; VEGF= vascular endothelial growth factor.

CRH regulates the hypothalamic pituitary adrenal (HPA) axis, which allows cells to respond to environmental stresses.⁸⁹ Melanocytes are known to secrete CRH, which permits them to produce cortisol and ACTH to maintain skin homeostasis.⁹⁰ In cell culture, melanoma cells treated with CRH migrated further during stress, which was determined to be mediated by the ERK1/2 pathway.⁹¹ Increases in the levels of CRH and proopiomelanocortin (POMC) are strongly associated with malignant melanoma.⁹² Another function of CRH in the skin is to act as a growth regulator by both promoting and suppressing cell proliferation. CRH receptor 1 (CRH-R1) controls the action of CRH and can promote cAMP and IP3 synthesis in dermal and epidermal cells.⁹³ Melanomas exclusively express CRH-R1 which plays an important role in proliferation and has become a agonist target.⁹⁴ In a B16 mouse melanoma model, daily injections with CRH were found to reduce tumor volume by 30–60% compared to control animals.⁹⁵ Research shows that CRH promotes survival of melanocytes during starvation and prevents cell proliferation by inhibiting growth factor signaling through cAMP and IP3 second messengers.⁹⁶ TNF-α, IL-1 and IL-6 stimulate HPA axis to produce CRH which drives a pro- inflammatory, Th1 biased response.⁹⁷ When peripheral CRH acts on the HPA axis in the brain, it triggers a classical feedback mechanism that leads to the secretion of cortisol, a steroid hormone with anti-inflammatory effects.⁹⁸ Thus, CRH is a potential modulator of a Th1 biased response and its role should be further studied in cancer.

Prolactin

Prolactin (PRL) is a polypeptide hormone secreted by the syncytiotrophoblast that reaches its highest levels during late pregnancy. (figure 4) PRL acts on corpus luteum cells in the ovary to stimulate secretion of progesterone to maintain pregnancy and on epithelial cells of the mammary gland to initiate milk production.⁹⁹ The amount of PRL secreted is directly proportional to the size of the fetus and it functions to provide energy for the mother and nutrients for the fetus. Through lipolysis and anti-insulin effects of prolactin action, the maternal insulin level increases, providing free fatty acids and amino acids to the growing fetus.¹⁰⁰ Prolactin functions to increase β-islet cell proliferation, inhibits apoptosis and causes β-islet cells to become more responsive to glucose.¹⁰¹ Glucocorticoid expression of Rasd1 near the end of gestation changes insulin secretion during pregnancy through the inhibitory effects of prolactin on Rasd1 transcription.¹⁰² Activation of dopamine neurons suppresses PRL secretion through placental lactogens.¹⁰³ The N-terminal 16K prolactin fragment had been determined to stop proliferation and migration of vascular endothelial cells and cause cell cycle arrest resulting in apoptosis.¹⁰⁴ High levels of 16K hPRL can be detected in serum and urine of women suffering from preeclampsia.¹⁰⁵

In melanoma, PRL drives Th1 biased immunity through the secretion of IL-1, IL-6, IL-10, IL-12 and IFN-γ by NK cells and B lymphocytes,^106,107 and plasma cell activation.¹⁰⁸ It increases T cell proliferation¹⁰⁹ and decreases B cell apoptosis.¹¹⁰ Through animal studies, prolactin’s anti-tumor effect has been shown to promote tumor specific macrophages through IFN-γ and IL-12, and CT26 tumor bearing mice injected with recombinant human PRL and IL-15 had enhanced cytotoxic activity to the tumor resulting in fewer lung metastasis and longer overall survival.^111,112 However, PRL does not have an anti-tumor effect in all cancers. In breast cancer, high levels of prolactin are associated with a greater risk for developing cancerous ERα+ tumors and is associated with much poorer outcomes for patients with these increased levels.¹¹³ This is likely due to the fact that PRL promotes mammary tumorigenesis independent of cyclin D1 activation.¹¹⁴ This has not yet been observed in melanoma. Pro-inflammatory cytokines such as IL-1, IL-2 and IL-6 produced by both pituitary and extrapituitary cells have a stimulatory effect on PRL secretion, while IFN-γ inhibits its production.¹¹⁵ Many immune cells express the prolactin receptor (PRLR) including monocytes, macrophages, B and T cells, granulocytes and NK cells.¹¹⁶ Prolactin secretion from the anterior pituitary gland is inhibited by the dopamine D2 receptor (D2R), which is widely expressed on melanoma cells and plays a key role in inhibiting adenylyl cyclase, which is necessary for cellular signal transduction.^117,118 PRL also enhances the effect of immune cells, including CD34+ stem cells, in the blood through upregulation of major histocompatibility complex (MHC) class II expression on antigen presenting cells, T cell clonal expansion, antibody production, increased cytotoxicity of NK cells and microbe killing of macrophages.¹¹⁹ In mouse melanoma models, 16K hPRL can block angiogenesis and inhibit tumor growth by activating NF-κB causing tumor infiltrating lymphocytes to access and destroy cancer cells.¹²⁰ This PRL peptide endogenously blocks Notch signaling which greatly impairs the tumor’s ability to vascularize.¹²¹ 16K hPRL has also been shown to block angiogenesis through inhibition of basic fibroblast growth factor and VEGF, making it a target for anti-tumor therapies.¹²² Thus, better understanding of how PRL can promote cell-mediated immunity could help researchers design better ways to initiate tumor destruction.

Visfatin

Previously known as pre-B-cell colony enhancing factor (PBEF) or nicotinamide phosphoribosyltransferase (Nampt), visfatin is a visceral fat cytokine with an important role in the promotion of inflammation.^123–125 Visfatin levels are highest at the end of gestation, and promote the secretion of IL-1β, IL-6, IL-8, COX-2, TNF-α and prostaglandin E₂.^26,127 Interestingly, infection associated preterm labor is concomitant with elevated visfatin circulating in maternal plasma.¹²⁸ These findings suggest visfatin could be important for immune resolution back to a Th1 state. However, the role of visfatin in melanoma is poorly understood. It appears that visfatin is more highly expressed in melanoma lesions compared to benign lesions.¹²⁹ Moreover, it seems that visfatin may promote melanoma cell growth in vitro. Namely, a study using melanoma Me45 cell line showed that visfatin increased proliferation of these tumor cells.¹³⁰ In another study, neutrophils have been found to synthesize visfatin in response to inflammatory stimulus and inhibit apoptosis in vitro and in sepsis patients.¹³¹ Thus, it remains possible that there could be an additional regulatory level of visfatin during pregnancy, which might be lost during tumorigenesis, and thereby promotes melanoma proliferation, but more studies need to be conducted to elucidate visfatin role in melanoma.

ESTROGEN: A DOUBLE-EDGED SWORD MODULATING TH1/TH2 IMMUNITY

Women have high levels of estrogen, which drop dramatically when they reach menopause. Estrogen modulates the immune response by inducing peripheral T cells to secrete pro-inflammatory cytokines IFN-γ and IL-2,¹³² but also promotes tolerance by inducing IL-10 secretion.¹³³ (figure 3 and 4) Estradiol-17β (E2) at high concentrations induces a Th1 response while at low concentrations it biases the system toward immune tolerance.¹³⁴ Pro-inflammatory responses are regulated by E2 through NF-κB. On the other hand, estrone (E3) is only detectable in pregnant women because it is produced by the placenta and fetus. E3 is important for reducing the production of pro-inflammatory cytokines in circulation by decreasing IκB degradation which inhibits NF-κB activation and apoptosis (figure 3).¹³⁵ Estrogens work through interaction with the estrogen receptor (ER) to induce transcriptional regulation. Immune cells including dendritic cells (DCs), natural killer (NK) cells, macrophages and lymphocytes express estrogen receptors, signifying that this hormone modulates their function.¹³⁶ Indeed, the differentiation of DCs is regulated by E2 acting upon ERα. E2 levels are the highest during the third trimester, when immature DCs are high and actively presenting fetal antigen to the mother’s immune system to begin labor.¹³⁷ Complement component 3 (C3), a protein that plays a central role in inducing inflammation in innate immunity, is increased by estrogen in oviductal epithelial cells.¹³⁸ Angiogenesis in the uterus is also driven by estrogen. VEGF mRNA levels increase in the endometrium in the presence of E2.¹³⁹ This increases placental blood flow and vasodilation which are both characteristic of angiogenesis. Cytotrophoblasts, but not syncytiotrophoblasts, stimulate VEGF production directly and this production is correlated to the levels of E2 in the serum.¹⁴⁰ Thus, estrogen stimulates VEGF production and blood vessel formation, which is essential for establishment and maintenance of the fetus during pregnancy.

Women at risk for familial breast cancer have increased risk for developing melanoma and vice versa, suggesting that estrogens can promote tumorigenesis.¹⁴¹ It has been known for many years that melanoma tumor cells express estrogen receptors.¹⁴² Estrogens affect lymphocytes by initiating the secretion of IL-10, IL-12 and IFN-γ and inhibition of TNF-α,¹⁴³ stimulation of antibody production,¹⁴⁴ and the reduction of macrophage and dendritic cell apoptosis.^145,146 E2 has been proven to inhibit melanoma growth by obstructing receptor binding of IL-8.⁷⁴ Chronic exposure to estrogens was suggested to increase NF-κB stimulation and induce pro-inflammatory responses by macrophages.¹⁴⁷ It has been shown that acute and chronic inflammation leads to the upregulation of ERα, but not ERβ, thus impacting the effects of estrogens on T cells, as T cells express ERα, while B cells express ERβ.¹⁴⁸ It was also reported that estrogen decreases apoptosis, as well as TNF-α production.¹⁴⁹ E2 and E3 each have a strong affinity for binding ERα,¹⁵⁰ which is why ERα is commonly associated with tumor promotion. Yet, ERβ has been found on malignant melanoma cells that are negative for ERα.¹⁵¹ Loss of ERβ expression correlated with increased invasiveness of the tumor,¹⁵² suggesting that loss of ERβ expression increases malignant transformation in melanoma. The most important prognostic factors in melanoma are tumor thickness (Breslow depth) and invasive level (Clark level); as each increase, the patient’s prognosis decreases.¹⁵³ Interestingly, increased ERβ expression is not found on non-malignant cells surrounding the tumor, only on the tumor cells themselves.¹⁵⁴ Another role of ERβ is regulation of monocyte apoptosis through the Fas/FasL signaling pathway.¹⁵⁵ It has been shown that myeloid progenitor cells exposed to GM-CSF differentiate into immature dendritic cells through expression of ERα and increased levels of estrogens.^156,157 Recently, 2–methoxyestradiol (2ME2), an estrogen derivative, has been determined to be a potent inhibitor of angiogenesis and melanoma growth in a mouse model.¹⁵⁸ Taken together, this indicates that estrogens can both promote or hinder tumor growth and monitoring ER expression could help clinicians determine the patient prognosis to the disease.

SEXUAL DIMORPHISM IN RESPONSE TO HORMONES

In this review, we have discussed existing data supporting a role for sex hormones and their ability to manipulate the immune system in both pregnancy and melanoma. Beyond these observations, several studies have identified sexual dimorphism with respect to inflammatory responses in a variety of settings. For example, sex-based differences in vascular function have been described and show the development of early atherosclerotic lesions and plaques with increased production of inflammatory mediators (IL-10 and TNF-α) in women compared to men.^159–161 That the state of pregnancy, and not a general state of tolerance in women, accounts for additional sex-based differences in immune function is supported by the fact that autoimmune diseases are more prevalent in women.¹⁶² Moreover, autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematous (SLE) are diagnosed more frequently in women and are strongly correlated with sex hormones.^163,164 Interestingly, some autoimmune diseases such as arthritis remit with pregnancy, but SLE has been found to often worsen in severity during gestation.¹⁶³ Women also exhibit enhanced innate immune responses to infections,^165,166 not blunted ones, and survive episodes of severe sepsis to a greater degree than men.¹⁶⁷ It is clear that sex plays a role in risk, severity and prognosis of many diseases, including cancer, and research in this area has just begun. It may one day become important to consider the sex of the cancer patient when personalizing cancer immunotherapies.

REVERSING TOLERANCE IN MELANOMA

In order to positively affect survival of patients with cancer, it appears critical to find a way to break immune tolerance to the malignancy. Insights into achieving this goal in metastatic melanoma may lie in well recognized abnormalities of pregnancy (e.g. spontaneous abortions). For example, serum taken from women suffering from recurrent spontaneous abortion have a cytokine profile indicative of a Th1 response compared to healthy pregnant women.¹⁶⁸ To counter this, intravenous immunoglobin (IVIG) given to women, with reported previous recurrent miscarriages and/or suffering from recurrent miscarriages after IVF, every 3–4 weeks starting from conception or 24 hours before embryo transfer showed a decrease of natural killer cells and a implantation success rate of 92.5%,compared to 25% for women not receiving IVIG.¹⁶⁹ Moreover, new immunotherapies for promoting tolerance to the fetus during gestation are being studied in animal models and could have important implications in melanoma. In mice, researchers have shown that blocking CD80 and CD86 enhances maternal tolerance decreasing implantation failure through the increase of Tregs and development of Th2 cells.¹⁷⁰ The use of anti-CTLA4 antibody therapy with CD80/CD86 blockade has been found to regulate the Th2/Th1 balance in PBMCs isolated from women with recurrent spontaneous abortion, which lead to the design of an adenoviral cytotoxic T lymphocyte antigen 4 (CTLA4) antibody that showed improved pregnancy outcomes in a mouse model of spontaneous abortion.^171,172

The opposite, immune activating effect is desired in metastatic melanoma. Recent FDA approval of the anti-CTLA4 antibody, ipilimumab, was developed to promote T cell activation and tumor destruction by promoting cytotoxic capacity of naturally occurring tumor specific T cells thereby overcoming, in part, their state of immune tolerance.¹⁷³ Metastatic melanoma patients treated with ipilimumab given alone or in combination with a peptide vaccine (gp100) exhibited a median overall survival time (OS) of approximately 10 months, an improvement over the 6.4 month OS observed in patients receiving gp100 vaccine alone.¹⁷⁴ As breakthrough as this drug is, it comes at a cost of significant immune related adverse events in up to 15% of patients.¹⁷⁵ Another new and promising antibody targets program death 1 (PD-1), an additional receptor found on activated cells which is critical for immune regulation, has been shown to promote a proinflammatory environment by IFN-γ and IL-2 secretion.¹⁷⁶ In clinical trials, this drug, following treatment with or without the anti-CTLA inhibitor, showed a high rate of sustained tumor regression and median overall survival of 11 months in patients with metastatic melanoma.¹⁷⁷ This lead to administering both ipilimumab and nivolumab (anti-PD-1) concurrently, resulting in 53% of patients obtaining an 80% reduction in tumor volume.¹⁷⁸ Unfortunately, the reported increase in objective response rate was paralleled with a similar increase in severe toxicity. Other promising immunotherapies for melanoma patients are also being studied. Anti-tumor immunity has been shown by utilizing an antibody to CCR4 a marker found on immune suppressive Treg cells, which when given to a T-cell leukemia-lymphoma patient resulted in a CD8+ T cell response to the tumor.¹⁷⁹ HER2/neu, an antibody commonly used to treat breast cancers, has recently shown positive effect on a number of melanoma cell lines and xenograft models.¹⁸⁰ Bevacizumab, an anti-VEGF antibody, given to metastatic melanoma patients in combination with albumin-bound paclitaxel and carboplatin, resulted in an increase of CD8+ lymphocytes but did not affect the Th1/Th2 ratio.¹⁸¹

Albeit promising, the clinical successes of present day immunotherapeutic strategies for metastatic cancer fail short of their preclinical results. Many share the belief that the reason for this discrepancy in clinical translation is the result of tumor driven immune tolerance of human cancer.¹⁸² Another reason for this discrepancy is that malignant melanocytes utilize melanogenesis, which is a normal metabolic process that generates a local immunosuppressive environment through POMC derived peptides and steroids.¹⁶ It was discovered that inhibiting melanogenesis increases the potency of the immune system and chemotherapy against tumors.¹⁸³ Overcoming the tumor induced modulation of systemic immunity in order to recover the ability of endogenously generated immune cells to effectively destroy the malignancy will be a significant challenge. An interesting concept in pregnancy, which could lead to better understanding of malignancy, is the spontaneous return to cytotoxicity near the end of parturition.¹³⁷ Near parturition, an unknown event results in the reactivation of Th1 maternal immunity and the initiation of labor (rejection). Characterizing this phenomenon, by further studying the role cytokines, cells and hormones play could translate into different methods for breaking tolerance in cancer patients, which could ultimately improve therapeutic efficacy and overall survival. The skin is a steroidogenic organ; it synthesizes its own steroids and sex hormones and can regulate local immune activities along with affecting the function of the epidermis.¹⁸⁴ CRH has been found to block human melanoma cell proliferation in vitro⁹⁶ and could provide additional therapeutic value when paired with a targeted agent. As progesterone promotes a tolerant state in pregnancy, treating a stable melanoma patient with progesterone was found to cause proliferation of dormant micrometastases by tipping the immune system back to a Th2 state.¹⁸⁵ B7-H1 (PD-L1), is a molecule expressed on antigen presenting cells and contributes to tumor evasion and expansion of Tregs.^186,187 In a B7-H1 knockout mouse model of melanoma, it was discovered that females had superior tumor growth resistance compared to males due to estrogen mediated suppression of Treg cells.¹⁸⁸ Clinically, in a phase I study of advanced cancers, including melanoma, patients treated with PD-1 antibody showed a significant response rate (28%) which was durable in only those patients expressing PD-L1, making it a potential biomarker for anti-PD-1 treatment response.¹⁸⁹ However, the authors did not compare men to women to look at potential differences due to sex. Altogether, in order to improve outcomes for melanoma patients treated with immunotherapeutics we must more completely understand the process of normal pregnancy immunoregulation, specifically the return to cytotoxicity (Th1 bias) at the end of gestation.

CONCLUSION

The dynamic maternal immune responses to normal pregnancy have evolved out of the need to support a semi-allogenic fetus over the duration of the pregnancy, without significant infectious or inflammatory impediment to mother . In turn, the maternal immune system is tightly regulated by hormone release and cytokine action to protect the developing fetus. Cancers, including melanoma, appear to induce similar tolerogenic immune programs through a variety of mechanisms, including paracrine secretion of sex hormones, to drive angiogenesis required for oxygen and nutrient supply, all while evading immune attack in a manner similar to the process of placentation. A better understanding of the molecular switches involved in the induction and reversal of immune tolerance in the setting of pregnancy may help identify new methods for targeted immune modulation for patients with melanoma.

Abbreviations

CRH

corticotrophin releasing hormone

IFN-γ

interferon gamma

IL

interleukin

DC

dendritic cells

E2

estradiol-17b

E3

estrone

ER

estrogen receptor

HCG

human chronic gonadotrophin

PIBF

progesterone induced blocking factor

PlGF

placental growth factor

PR

progesterone receptor

PRL

prolactin

Treg

T regulatory cell

uNK

uterine natural killer cell

VEGF

vascular endothelial growth factor