- Email: [email protected]
Gonadotropin stimulation in mice leads to ovarian accumulation of immature myeloid cells and altered expression of proangiogenic genes
European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Gonadotropin stimulation in mice leads to ovarian accumulation of immature myeloid cells and altered expression of proangiogenic genes N. Pencovich a,b,c , S. Hantisteanu a,b , M. Hallak a , O. Fainaru a,c,d, * a Laboratory for Reproductive Immunology, Department of Obstetrics and Gynaecology, Hillel Yaffe Medical Centre, Faculty of Medicine, Technion, Israel Institute of Technology, Hadera, Israel b Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel c Department of Surgery, Tel Aviv Souraski Medical Center, Tel Aviv, Israel d IVF Unit, Department of Obstetrics and Gynaecology, Hillel Yaffe Medical Centre, Faculty of Medicine, Technion, Israel Institute of Technology, Hadera, Israel
A R T I C L E I N F O
A B S T R A C T
Article history: Received 28 September 2013 Received in revised form 25 March 2014 Accepted 22 May 2014
Objective: Ovarian hyperstimulation syndrome is associated with increased angiogenesis and vascular leakage. Immature myeloid cells (IMCs) and dendritic cells have been shown to be actively involved in angiogenesis in several disease models in mice and humans. Nevertheless, little is known about the role of these cells in the ovary. As such, this study sought to determine whether alterations in these ovarian myeloid cell populations are associated with gonadotropin stimulation in a mouse model. Study design: Four-week-old pre-pubertal C57Bl/6 female mice were allocated into three groups: highdose stimulation (n = 4; pregnant mare serum gonadotropins (PMSG) 20 U for 2 days), low-dose stimulation (n = 5; PMSG 5 U for 1 day) and sham-treated controls (n = 4). Human chorionic gonadotropin 5 U was injected on Day 3, and the mice were killed on Day 5. Ovaries were analysed by flow cytometry, confocal microscopy and quantitative polymerase chain reaction. Results: Gonadotropin stimulation increased the proportion of CD11b+Gr1+ IMCs among the ovarian myeloid cells: 22.6 8.1% (high dose), 7.2 1.6% (low dose) and 4.1 0.3% (control) (p = 0.02). Conversely, gonadotropin stimulation decreased the proportion of ovarian CD11c+MHCII+ dendritic cells: 15.1 1.9% (high dose), 20.7 4.8% (low dose) and 27.3 8.2% (control) (p = 0.02). IMCs, unlike dendritic cells, were localized adjacent to PECAM1+ endothelial cells. Finally, gonadotropin stimulation was associated with increased expression of S100A8, S100A9, Vcan and Dmbt1, and decreased expression of MMP12. Conclusions: Gonadotropin stimulation is associated with proangiogenic myeloid cell alterations, reflected by a dose-dependent increase in ovarian IMCs and a parallel decrease in dendritic cells. Recruited IMCs localize strategically at sites of angiogenesis. These changes are associated with differential expression of key proangiogenic genes. ã 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Immature myeloid cells Dendritic cells Ovarian stimulation Angiogenesis
Introduction Ovarian hyperstimulation syndrome (OHSS) is a rare iatrogenic disorder that may occur when heavily stimulated ovaries, containing numerous follicles with high oestradiol production, are exposed to exogenous human chorionic gonadotropin (hCG) [1,2]. OHSS is characterized by ovarian enlargement, increased vascular permeabilityand fluid shifting into the third space. This results in abdominal distension and increased intra-abdominal pressure concomitant with haemoconcentration [3]. The pathogenesis of OHSS is unclear.
* Corresponding author at: IVF Unit and Laboratory for Reproductive Immunology, Department of Obstetrics and Gynaecology, Hillel Yaffe Medical Centre, P.O. Box 169, Hadera 38100, Israel. Tel.: +972 46188405; fax: +972 46188406. E-mail addresses: [email protected], [email protected] (O. Fainaru). http://dx.doi.org/10.1016/j.ejogrb.2014.05.025 0301-2115/ ã 2014 Elsevier Ireland Ltd. All rights reserved.
However, several factors that are induced by exogenous hCG and affect vascular permeability and angiogenesis, such as vascular endothelial growth factor (VEGF), transforming growth factor, platelet-derived growth factor, angiopoietins, prostaglandins and others, have been shown to play a role in this acute process [4–9]. Immature myeloid cells (IMCs), also known as myeloid-derived suppressor cells, have been shown to play an active role in several biological processes that involve angiogenesis. These cells have been shown to infiltrate placentas of pregnant mice and humans, and actively promote angiogenesis. Their presence is correlated with placental weight and birth weight [10,11]. IMCs have been shown to promote tumour growth and metastasis by modulating the cytokine environment and promoting angiogenesis [12–14]. Dendritic cells are specialized antigen-presenting cells that play key roles in the initiation and modulation of the adaptive immune response [15]. Tumour-infiltrating dendritic cells have been shown
76
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
to promote tumour growth and angiogenesis by secreting proangiogenic molecules, assembling into tumour neovessels, and/or transdifferentiating into endothelial-like cells [16–18]. The authors have recently shown that with striking similarity to their role in tumour growth and angiogenesis, supplementation of dendritic
cells augmented the growth of endometriosis lesions and intralesion angiogenesis [19]. IMCs and dendritic cells are at different stages of myeloid development, as IMCs represent cells in early phases of development while dendritic cells are mature, fully differentiated cells.
Fig. 1. Mouse model of ovarian stimulation. (A) Representative images of harvested murine ovaries 48 h after human chorionic gonadotropin, following intraperitoneal administration of low and high doses of pregnant mare serum gonadotropins (PMSG). (B) Representative images of haematoxylin and eosin staining of 10 mm histological sections of a control ovary vs a high-dose hyperstimulated ovary. Images were viewed with a Nikon Eclipse 800 microscope, captured using a Nikon DXM1200 digital camera, and processed with Nikon ACT-1 2.63 software. (C) Bar histograms depicting average gross weights of mice (right) and ovaries (two per mouse) (left) harvested from control mice (cont, n = 4), low-dose hyperstimulated mice (low, n = 5) and high-dose hyperstimulated mice (high, n = 4). *p < 0.05 (Mann–Whitney test).
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Under normal physiological conditions, IMCs that are generated in the bone marrow differentiate into mature granulocytes, macrophages or dendritic cells [20]. However, under pathological conditions such as cancer, a partial block in this differentiation process leads to expansion of the IMC population. The goal of this study was to investigate whether alterations in ovarian myeloid cell populations (i.e. IMCs and dendritic cells) are associated with increased gonadotropin stimulation in mice.
77
Technologies). Quantitative polymerase chain reaction (PCR) was performed using a light cycler 480 (Roche, Nutley, NJ, USA) with 480 SYBR Green I Master (Roche, Nutley, NJ, USA) and the following PCR
Materials and methods Experimental Animal studies were undertaken using four-week-old prepubertal female mice. C57Bl/6J mice were purchased from Harlan Laboratories (Jerusalem, Israel). The murine OHSS model was performed as described previously [21]. Briefly, pregnant mare serum gonadotropin (Sigma–Aldrich) was administered intraperitoneally as a high dose (20 U PMSG for 2 days), low dose (5 U PMSG for 1 day) or sham (control) injection (0 U PMSG). hCG 5 U was injected on Day 3 to induce ovulation. Mice were killed 48 h after hCG administration, and the ovaries were harvested. All animal procedures were performed in compliance with the guidelines of Weizmann Institute of Science, and protocols were approved by the Institutional Animal Care and Use Committee. Flow cytometry Ovarian specimens were digested with an enzyme mixture including DNase-I 100 mg/ml (Roche, Mannheim, Germany) and collagenase type IV 1 mg/ml (Sigma–Aldrich, Israel), dissolved in phosphate-buffered saline (PBS), at 37 C for 30 min. Digested tissue was filtered through a 40 mM cell strainer and resuspended in fluorescence-activated cell-sorting buffer (PBS, EDTA5 mM and 3% fetal calf serum). Immunostaining was performed in the presence of rat anti-mouse Fc-gamma receptor III/II (CD16/32; eBioscience, San Diego, CA, USA) by incubating the cells with monoclonal antibodies for 30 min on ice. Staining reagents included fluorochrome (APC, PE or FITC)-labelled anti-CD11b, CD45, Gr1, CD11c and MHC-II (all purchased from eBioscience). Flow cytometry was performed with a FACSORT (Becton Dickinson, Mountain View, CA, USA) and analysed using Flowjo Software (Tree Star Inc., Ashland, OR, USA). IMCs were defined as CD45+CD11b+Gr1+ cells and dendritic cells were defined as CD45 + CD11c+MHCIIhigh cells. Immunohistochemistry Frozen sections (12 mm) of ovarian tissue were fixed with acetone ( 20 C), washed in PBS and blocked with 20% horse serum in PBS. Immunostaining was performed with rat antimouse PECAM1 PE (BD Biosciences Pharmingen), biotin antimouse CD11b and biotin anti-mouse CD11c (BD Biosciences Pharmingen). Streptavidin-cy2 (Jackson ImmunoResearch, Suffolk, UK) was used as the secondary antibody for the CD11b and CD11c staining. Sections were mounted on glass slides with ProLong Gold anti-fade reagent and DAPI mounting media (Invitrogen Life Technologies, Burlington, Ontario, Canada). Ovarian morphology and the presence of corpura lutea were determined in sections stained with haematoxylin and eosin. RNA processing and real-time polymerase chain reaction RNA was isolated by EZ-RNA (Biological Industries, Beit Haemek, Israel), according to the manufacturer's instructions. CDNA was amplified with SuperScript II reverse transcriptase (Invitrogen Life
Fig. 2. Recruitment of bone-marrow-derived CD45+ cells into stimulated ovaries. (A) Representative flow cytometry histograms of anti-CD45 stained single cell suspensions, derived from control ovaries (top graph, cont, n = 3), low-dose stimulated ovaries (middle graph, low, n = 5) and high-dose stimulated ovaries (bottom graph, high, n = 4) (FSC: forward scatter). (B) Bar histogram depicts the percentages of CD45+ cells among the ovarian cells described in (A). Data represent mean standard deviation. *p < 0.05 (Mann–Whitney test).
78
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
primers: S100a8, F-AAATCACCATGCCCTCTACAAG, R-CCCACTTTTATCACCATCGCAA; S100a9, F-ATACTCTAGGAAGGAAGGACACC, R-TCCATGATGTCATTTATGAGGGC; Dmbt1, F-GAGGACTATCTGTGGAGATG, R-TTTGTCCCTCCTGGATTCCA; Vcan, F-TGGGATTGAAGACACTCAGG, R-ATCAGACAGCCATCCTGCAT; Mmp12, F-CTGCTCCCATGAATGACAGTG, R-AGTTGCTTCTAGCCCAAAGAAC; and Actb (used for normalizing control), F-GGCTGTATTCCCCTCCATCG, R-CCAGTTGGTAACAATGCCATGT.
Statistical analysis Data presented in the histograms are expressed as mean standard deviation. Significance of the differences between continuous variables was determined using the two-tailed Mann–Whitney non-parametric test. Real-time PCR expression data were analysed using the unpaired two-tailed Student's t-test. p < 0.05 was considered to indicate significance.
Fig. 3. Ovarian stimulation results in enrichment of immature myeloid cells (IMCs) and depletion of dendritic cells (DCs). Single cell suspensions, derived from control (n = 3), low-dose stimulated ovaries (n = 5) and high-dose stimulated ovaries (n = 4) were analysed by flow cytometry. (A) Representative images of the proportion of CD11b+Gr1+ IMCs among the CD45+ cells (gating as in Fig. 2A) in control ovaries (top graph), low-dose hyperstimulated ovaries (middle graph) and high-dose hyperstimulated ovaries (bottom graph). Bar histogram depicts the percentages of CD11b+Gr1+ IMCs within ovaries. (B) Representative images of the proportion of CD11c+MHCIIhigh dendritic cells among the CD45+ cells in control ovaries (top graph), low-dose hyperstimulated ovaries (middle graph) and high-dose hyperstimulated ovaries (bottom graph). Bar histogram depicts the percentages of CD11c+MHCII+ dendritic cells within ovaries. Data represent mean standard deviation. *p < 0.05 (Mann–Whitney test).
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Results Ovarian stimulation leads to a dose-dependent enrichment of ovarian IMCs paralleled by a reciprocal reduction of ovarian dendritic cells The effect of gonadotropin stimulation on gross weight, ovarian weight and histological appearance was examined. Mice in both stimulated groups demonstrated a slight, yet significant, increase in weight: 14.2 0.8 g (control), 16.1 1.3 g (low dose) and 15.2 0.9 g (high dose) (p < 0.05). Ovaries harvested 48 h after hCG treatment in mice that were treated with low-dose (5 U PMSG) and high-dose (20 U PMSG) gonadotropins were significantly larger and demonstrated more than a four-fold increase in weight compared with sham-treated control ovaries (Fig. 1A, C): 4.4 2.4 mg (control), 22.1 0.8 mg (low dose) and 21.7 5.8 g (high dose) (p < 0.05). Multiple large follicles and numerous postovulatory corpora lutea were identified (Fig. 1B). No significant differences in weight and general appearance of the ovaries were noticed between low- and high-dose groups (Fig. 1A, C). Analysis of ovarian cells by flow cytometry demonstrated a small, yet significant, increase in the ovarian CD45+ bonemarrow-derived cell population in both low- and high-dose groups compared with the control group (Fig. 2): 1.6 0.7% (control), 4.5 1.2% (low dose) and 3.0 1.7% (high dose) (p < 0.05). Importantly, upon ovarian stimulation, the proportion of CD11b+Gr1+ IMCs among the CD45+ bone-marrow-derived cells in the ovaries was higher compared with the control ovaries. Of note, this increase in the ovarian CD11b+Gr1+ IMC population was
79
dose-dependent, as approximately a two-fold increase and a fivefold increase were observed following low- and high-dose gonadotropin treatment, respectively (Fig. 3A): 4.1 0.3% (control), 7.2 1.6% (low dose) and 22.6 8.1% (high dose) (p < 0.05). Concomitantly, a dose-dependent reduction in the proportion of CD11c+MHCII+ dendritic cells among the CD45+ bone-marrowderived cells was noted, as approximately a two-fold and a threefold decrease in this cell population was demonstrated following low- and high-dose gonadotropin treatment, respectively (Fig. 3B): 27.3 8.2% (control), 20.7 4.8% (low dose) and 15.1 1.9% (high dose) (p < 0.05). This finding was also confirmed by immunohistochemistry, with a marked increase in IMCs paralleled by a decrease in dendritic cells upon ovarian stimulation (Fig. 4A, B). Of note, upon ovarian stimulation, IMCs seem to be recruited to an area adjacent to the follicular wall (Fig. 4A), whereas in unstimulated ovaries, dendritic cells are localized sparsely within the stroma (Fig. 4B). Ovarian stimulation is associated with differential expression of key angiogenic genes The global expression pattern of IMCs has been characterized previously in both normal and malignant processes in which these cells are required to support angiogenesis [11,22,23]. This study examined the expression of key genes that have previously been shown to play roles in angiogenesis, and were regulated differentially in IMCs from dormant vs angiogenic tumours, or tumours vs placentas [24–31]. The results indicated that the
Fig. 4. Immature myeloid cells (IMCs) are recruited to the follicular wall upon ovarian stimulation with pregnant mare serum gonadotropins (PMSG). Frozen sections of control and high-dose stimulated ovarian tissues were immunostained with anti-PECAM1 antibody (orange), anti-CD11b (IMCs, green) (A) and anti-CD11c (dendritic cells, green) (B). 4'-6-Diamidino-2-phenylindole (DAPI) (blue) was used to detect nuclei.
80
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Fig. 5. Differential gene expression in control (WT: wild type) vs high-dose stimulated ovaries (ovarian hyperstimulation syndrome, OHSS). Quantitative reverse transcriptase polymerase chain reaction analysis of relative expression of the indicated genes in control vs high-dose hyperstimulated ovaries. Data represent mean standard deviation of experiments performed in triplicate. *The differences in upregulated genes presented in (A) and downregulated genes presented in (B) were significant (Student's t-test, p < 0.05). Genes presented in (C) did not demonstrate a significant change in expression between the control and high-dose hyperstimulated ovaries.
expression of S100a8, S100a9, Dmbt1 and Vcan was significantly upregulated in stimulated ovaries compared with control ovaries (Fig. 5A). On the other hand, the expression of Mmp12 was significantly downregulated in stimulated ovaries compared with control ovaries (Fig. 5B). Of note, the expression of other genes that are differentially expressed in different IMC populations and known to play key roles in angiogenesis, including Il1b, Jag1, Twist1, Tgfbr3, Klk1, Serpine1 and Ppbp, did not change significantly upon ovarian hyperstimulation (Fig. 5C). Comment The central feature of OHSS is the development of vascular hyperpermeability. As hCG causes the ovary to undergo extensive luteinization, large amounts of oestradiol and progesterone are released, along with other local cytokines. Various vasoactive substances such as VEGF, tumour necrosis factor-a, endothelin-1 and others are secreted by the multiple corpora lutea [3,32]. These, in turn, lead to vascular hyperpermeability by making local
capillaries ‘leaky’, leading to a shift of fluids from the intravascular system to the abdominal and pleural cavities. Nevertheless, the cellular mechanisms that govern this complex process (i.e. which specific cells within the ovary release these vasoactive substances) is still largely unknown. This study used a murine model to examine the alterations in myeloid cell populations within the ovaries upon ovarian stimulation. Both low- and high-dose stimulation were used to control for dose-dependent effects. Both low- and high-dose stimulation led to a similar marked increase in ovarian size and weight, and the appearance of multiple mature follicles and corpora lutea. Of note, the increase in total body weight of the mice was small. It is possible that a more significant increase in weight would have been demonstrated if higher doses of gonadotropins were administered. However, this was not done in the present study in order to reduce unnecessary animal suffering. Upon ovarian stimulation, a small, yet significant, increase in the percentage of CD45+ bone-marrow-derived cells in the ovaries
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
was noted. These cells represent the total haematopoietic cell compartment within the ovaries, and include different cell populations including IMCs and dendritic cells. This mild increase may represent a net influx of blood cells into the ovaries during ovarian stimulation. Notably, this increase was observed in mice stimulated with both low- and high-dose gonadotropins. Although there was a slight trend for an increase in the low-dose group compared with the high-dose group, this trend was insignificant and likely to be of no biological significance. The striking dose-dependent increase in ovarian CD11b+Gr1+ IMCs in parallel to the decrease in CD11c+MHC-II+ dendritic cells upon stimulation suggests an intriguing notion that a shift in myeloid populations at different levels of maturation may affect ovarian angiogenesis, blood vessel permeability and, as a result, the severity of OHSS. IMCs populate the unstimulated ovary in very low numbers, while gonadotropin stimulation leads to a more than five-fold increase in this cell population. This signifies the importance of these cells in this process, presumably as regulators of angiogenesis and increased vascular permeability. The fact that IMCs localize to the vicinity of the follicular wall, a site known for intense angiogenesis within the ovary, suggests that IMCs in the ovary also play an active role in angiogenesis, as described in pregnancy [10,11] and in cancer [18,23]. Additionally, the differential location of dendritic cells (sparsely within the stroma) and IMCs (follicular wall) further implies that these are different cells with different physiological roles within the ovary. Previous studies [18,19,33,34] have shown that angiogenesis in general (as demonstrated in Matrigel plugs in vivo) and angiogenesis observed in tumours, endometriosis and choroidal neovascularization is dependent on the presence of immature dendritic cells. Specifically, supplementation of immature dendritic cells, but not mature dendritic cells, enhanced lesion growth. Depletion of dendritic cells in a transgenic mouse model that allows for their conditional ablation completely abrogated angiogenesis in Matrigel plugs, decreased tumour growth and decreased the growth of endometriotic lesions. Accordingly, the present study demonstrated that increased angiogenesis in the ovary is associated with ovarian stimulation, and ovulation is accompanied by a decrease in non-angiogenic mature dendritic cells. Nevertheless, the presence of proangiogenic immature dendritic cells in the ovary was not detected. Interestingly, the authors have recently demonstrated [35] that labour and delivery are preceded by similar myeloid cell alterations, reflected by a decrease in IMCs and an increase in dendritic cells populating the mouse placenta. It is therefore tempting to speculate that a governing mechanism applies both in midtrimester placentas and stimulated ovaries, involving the presence of IMCs and immature dendritic cells when active angiogenesis takes place. It is possible that when angiogenesis is no longer needed, IMCs mature into more differentiated cells such as dendritic cells. Increased expression of key genes, shown to be highly expressed by IMCs and to play a role in angiogenesis, further supports the notion that these cells contribute to vascular permeability in the pathogenesis of OHSS. S100a8 and S100a9, known to be involved in angiogenesis [36], were both highly expressed in placental- and tumour-derived IMCs [23]. Dmbt1 and Vcan, also key players in angiogenesis [25,26], were highly expressed in tumour-derived IMCs [22]. Although the global gene expression pattern of ovarian IMCs has not been described to date, the fact that Mmp12 is highly expressed by IMCs in other systems and downregulated upon ovarian stimulation, and that other key IMC genes such as Il1b, Jag1, Twist1, Tgfbr3, Klk1, Serpine1 and Ppbp were unaffected suggests that the ovarian IMC gene expression pattern is unique. Overall, these data suggest that changes in proangiogenic myeloid populations within the ovary may play a role in ovarian
81
hyperstimulation. Key angiogenic factors are expressed within recruited IMCs, and may influence vascular permeability and angiogenesis. Studies testing therapies that target these different myeloid cell populations as a novel strategy for the prevention and treatment of OHSS in humans are warranted. References [1] Avecillas JF, Falcone T, Arroliga AC. Ovarian hyperstimulation syndrome. Crit Care Clin 2004;20:679–95 ix. [2] Rydberg E, Pedersen-Bjergaard K. Effect of serum gonadotropin and chorionic gonadotropin on the human ovary. JAMA 1943;121:1117–22. [3] Grossman LC, Michalakis KG, Browne H, Payson MD, Segars JH. The pathophysiology of ovarian hyperstimulation syndrome: an unrecognized compartment syndrome. Fertil Steril 2010;94:1392–8. [4] Levin ER, Rosen GF, Cassidenti DL, et al. Role of vascular endothelial cell growth factor in ovarian hyperstimulation syndrome. J Clin Invest 1998;102:1978–85. [5] Navot D, Margalioth EJ, Laufer N, et al. Direct correlation between plasma renin activity and severity of the ovarian hyperstimulation syndrome. Fertil Steril 1987;48:57–61. [6] Garcia-Velasco JA, Pellicer A. New concepts in the understanding of the ovarian hyperstimulation syndrome. Curr Opin Obstet Gynecol 2003;15:251–6. [7] Scotti L, Abramovich D, Pascuali N, et al. Involvement of the ANGPTs/Tie-2 system in ovarian hyperstimulation syndrome (OHSS). Mol Cell Endocrinol 2012;365:223–30. [8] Soares SR. Etiology of OHSS and use of dopamine agonists. Fertil Steril 2012;97:517–22. [9] Budev MM, Arroliga AC, Falcone T. Ovarian hyperstimulation syndrome. Crit Care Med 2005;33(Suppl. 10):S301–6. [10] Mei-Dan E, Fainaru O, Hantisteanu S, Shavit T, Ellenbogen A, Hallak M. Proangiogenic immature myeloid cells populate the human placenta and their presence correlates with placental and birthweight. Am J Obstet Gynecol 2012;207:141e1–5. [11] Fainaru O, Hantisteanu S, Hallak M. Immature myeloid cells accumulate in mouse placenta and promote angiogenesis. Am J Obstet Gynecol 2011;204:544e18–23. [12] Yang L, DeBusk LM, Fukuda K, et al. Expansion of myeloid immune suppressor Gr+CD11b+ cells in tumor-bearing host directly promotes tumor angiogenesis. Cancer Cell 2004;6:409–21. [13] Yan HH, Pickup M, Pang Y, et al. Gr1+CD11b+ myeloid cells tip the balance of immune protection to tumor promotion in the premetastatic lung. Cancer Res 2010;70:6139–49. [14] Shojaei F, Wu X, Malik AK, et al. Tumor refractoriness to anti-VEGF treatment is mediated by CD11b+Gr1+ myeloid cells. Nat Biotechnol 2007;25:911–20. [15] Banchereau J, Steinman RM. Dendritic cells and the control of immunity. Nature 1998;392:245–52. [16] Sozzani S, Rusnati M, Riboldi E, Mitola S, Presta M. Dendritic cell–endothelial cell cross-talk in angiogenesis. Trends Immunol 2007;28:385–92. [17] Murdoch C, Muthana M, Coffelt SB, Lewis CE. The role of myeloid cells in the promotion of tumour angiogenesis. Nat Rev Cancer 2008;8:618–31. [18] Fainaru O, Almog N, Yung CW, et al. Tumor growth and angiogenesis are dependent on the presence of immature dendritic cells. FASEB J 2010;24:1411–8. [19] Fainaru O, Adini A, Benny O, et al. Dendritic cells support angiogenesis and promote lesion growth in a murine model of endometriosis. FASEB J 2008;22:522–9. [20] Satpathy AT, Wu X, Albring JC, Murphy KM. Re(de)fining the dendritic cell lineage. Nat Immunol 2012;13:1145–54. [21] Fainaru O, Hornstein MD, Folkman J. Doxycycline inhibits vascular leakage and prevents ovarian hyperstimulation syndrome in a murine model. Fertil Steril 2009;92:1701–5. [22] Pencovich N, Hantisteanu S, Wurtzel O, Hallak M, Fainaru O. Unique expression patterns associated with preferential recruitment of immature myeloid cells into angiogenic versus dormant tumors. Genes Immun 2012;14:90–8. [23] Fainaru O, Pencovich N, Hantisteanu S, Yona G, Hallak M. Immature myeloid cells derived from mouse placentas and malignant tumors demonstrate similar proangiogenic transcriptional signatures. Fertil Steril 2013;99:910–7. [24] Chen P, Huang Y, Bong R, et al. Tumor-associated macrophages promote angiogenesis and melanoma growth via adrenomedullin in a paracrine and autocrine manner. Clin Cancer Res 2011;17:7230–9. [25] Muller H, Hu J, Popp R, et al. Deleted in malignant brain tumors 1 is present in the vascular extracellular matrix and promotes angiogenesis. Arterioscler Thromb Vasc Biol 2012;32:442–8. [26] Vazquez-Ortiz G, Pina-Sanchez P, Vazquez K, et al. Overexpression of cathepsin F, matrix metalloproteinases 11 and 12 in cervical cancer. BMC Cancer 2005;5:68. [27] Min JK, Park H, Choi HJ, et al. The WNT antagonist dickkopf2 promotes angiogenesis in rodent and human endothelial cells. J Clin Invest 2011;121:1882–93. [28] Cheng P, Jiang FH, Zhao LM, et al. Human macrophage metalloelastase correlates with angiogenesis and prognosis of gastric carcinoma. Dig Dis Sci 2010;55:3138–46. [29] Hu L, Roth JM, Brooks P, Ibrahim S, Karpatkin S. Twist is required for thrombininduced tumor angiogenesis and growth. Cancer Res 2008;68:4296–302.
82
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
[30] Sahin H, Borkham-Kamphorst E, Kuppe C, et al. Chemokine Cxcl9 attenuates liver fibrosis-associated angiogenesis in mice. Hepatology 2012;55:1610–9. [31] Dong M, How T, Kirkbride KC, et al. The type III TGF-beta receptor suppresses breast cancer progression. J Clin Invest 2007;117:206–17. [32] Kumar P, Sait SF, Sharma A, Kumar M. Ovarian hyperstimulation syndrome. J Hum Reprod Sci 2011;4:70–5. [33] Pencovich N, Luk J, Hantisteanu S, Hornstein MD, Fainaru O. The development of endometriosis in a murine model is dependent on the presence of dendritic cells. Reprod Biomed Online 2014;28:515–21.
[34] Nakai K, Fainaru O, Bazinet L, et al. Dendritic cells augment choroidal neovascularization. Investig Ophthalmol Vis Sci 2008;49:3666–70. [35] Fainaru O, Paz G, Hantisteanu S, Hallak M. Myeloid cell alterations in the mouse placenta precede the onset of labor and delivery. Am J Obstet Gynecol 2014;210:73e1–7. [36] Li C, Li S, Jia C, Yang L, Song Z, Wang Y. Low concentration of S100A8/9 promotes angiogenesis-related activity of vascular endothelial cells: bridges among inflammation, angiogenesis, and tumorigenesis? Mediators Inflamm 2012;2012:248574.
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Gonadotropin stimulation in mice leads to ovarian accumulation of immature myeloid cells and altered expression of proangiogenic genes N. Pencovich a,b,c , S. Hantisteanu a,b , M. Hallak a , O. Fainaru a,c,d, * a Laboratory for Reproductive Immunology, Department of Obstetrics and Gynaecology, Hillel Yaffe Medical Centre, Faculty of Medicine, Technion, Israel Institute of Technology, Hadera, Israel b Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel c Department of Surgery, Tel Aviv Souraski Medical Center, Tel Aviv, Israel d IVF Unit, Department of Obstetrics and Gynaecology, Hillel Yaffe Medical Centre, Faculty of Medicine, Technion, Israel Institute of Technology, Hadera, Israel
A R T I C L E I N F O
A B S T R A C T
Article history: Received 28 September 2013 Received in revised form 25 March 2014 Accepted 22 May 2014
Objective: Ovarian hyperstimulation syndrome is associated with increased angiogenesis and vascular leakage. Immature myeloid cells (IMCs) and dendritic cells have been shown to be actively involved in angiogenesis in several disease models in mice and humans. Nevertheless, little is known about the role of these cells in the ovary. As such, this study sought to determine whether alterations in these ovarian myeloid cell populations are associated with gonadotropin stimulation in a mouse model. Study design: Four-week-old pre-pubertal C57Bl/6 female mice were allocated into three groups: highdose stimulation (n = 4; pregnant mare serum gonadotropins (PMSG) 20 U for 2 days), low-dose stimulation (n = 5; PMSG 5 U for 1 day) and sham-treated controls (n = 4). Human chorionic gonadotropin 5 U was injected on Day 3, and the mice were killed on Day 5. Ovaries were analysed by flow cytometry, confocal microscopy and quantitative polymerase chain reaction. Results: Gonadotropin stimulation increased the proportion of CD11b+Gr1+ IMCs among the ovarian myeloid cells: 22.6 8.1% (high dose), 7.2 1.6% (low dose) and 4.1 0.3% (control) (p = 0.02). Conversely, gonadotropin stimulation decreased the proportion of ovarian CD11c+MHCII+ dendritic cells: 15.1 1.9% (high dose), 20.7 4.8% (low dose) and 27.3 8.2% (control) (p = 0.02). IMCs, unlike dendritic cells, were localized adjacent to PECAM1+ endothelial cells. Finally, gonadotropin stimulation was associated with increased expression of S100A8, S100A9, Vcan and Dmbt1, and decreased expression of MMP12. Conclusions: Gonadotropin stimulation is associated with proangiogenic myeloid cell alterations, reflected by a dose-dependent increase in ovarian IMCs and a parallel decrease in dendritic cells. Recruited IMCs localize strategically at sites of angiogenesis. These changes are associated with differential expression of key proangiogenic genes. ã 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Immature myeloid cells Dendritic cells Ovarian stimulation Angiogenesis
Introduction Ovarian hyperstimulation syndrome (OHSS) is a rare iatrogenic disorder that may occur when heavily stimulated ovaries, containing numerous follicles with high oestradiol production, are exposed to exogenous human chorionic gonadotropin (hCG) [1,2]. OHSS is characterized by ovarian enlargement, increased vascular permeabilityand fluid shifting into the third space. This results in abdominal distension and increased intra-abdominal pressure concomitant with haemoconcentration [3]. The pathogenesis of OHSS is unclear.
* Corresponding author at: IVF Unit and Laboratory for Reproductive Immunology, Department of Obstetrics and Gynaecology, Hillel Yaffe Medical Centre, P.O. Box 169, Hadera 38100, Israel. Tel.: +972 46188405; fax: +972 46188406. E-mail addresses: [email protected], [email protected] (O. Fainaru). http://dx.doi.org/10.1016/j.ejogrb.2014.05.025 0301-2115/ ã 2014 Elsevier Ireland Ltd. All rights reserved.
However, several factors that are induced by exogenous hCG and affect vascular permeability and angiogenesis, such as vascular endothelial growth factor (VEGF), transforming growth factor, platelet-derived growth factor, angiopoietins, prostaglandins and others, have been shown to play a role in this acute process [4–9]. Immature myeloid cells (IMCs), also known as myeloid-derived suppressor cells, have been shown to play an active role in several biological processes that involve angiogenesis. These cells have been shown to infiltrate placentas of pregnant mice and humans, and actively promote angiogenesis. Their presence is correlated with placental weight and birth weight [10,11]. IMCs have been shown to promote tumour growth and metastasis by modulating the cytokine environment and promoting angiogenesis [12–14]. Dendritic cells are specialized antigen-presenting cells that play key roles in the initiation and modulation of the adaptive immune response [15]. Tumour-infiltrating dendritic cells have been shown
76
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
to promote tumour growth and angiogenesis by secreting proangiogenic molecules, assembling into tumour neovessels, and/or transdifferentiating into endothelial-like cells [16–18]. The authors have recently shown that with striking similarity to their role in tumour growth and angiogenesis, supplementation of dendritic
cells augmented the growth of endometriosis lesions and intralesion angiogenesis [19]. IMCs and dendritic cells are at different stages of myeloid development, as IMCs represent cells in early phases of development while dendritic cells are mature, fully differentiated cells.
Fig. 1. Mouse model of ovarian stimulation. (A) Representative images of harvested murine ovaries 48 h after human chorionic gonadotropin, following intraperitoneal administration of low and high doses of pregnant mare serum gonadotropins (PMSG). (B) Representative images of haematoxylin and eosin staining of 10 mm histological sections of a control ovary vs a high-dose hyperstimulated ovary. Images were viewed with a Nikon Eclipse 800 microscope, captured using a Nikon DXM1200 digital camera, and processed with Nikon ACT-1 2.63 software. (C) Bar histograms depicting average gross weights of mice (right) and ovaries (two per mouse) (left) harvested from control mice (cont, n = 4), low-dose hyperstimulated mice (low, n = 5) and high-dose hyperstimulated mice (high, n = 4). *p < 0.05 (Mann–Whitney test).
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Under normal physiological conditions, IMCs that are generated in the bone marrow differentiate into mature granulocytes, macrophages or dendritic cells [20]. However, under pathological conditions such as cancer, a partial block in this differentiation process leads to expansion of the IMC population. The goal of this study was to investigate whether alterations in ovarian myeloid cell populations (i.e. IMCs and dendritic cells) are associated with increased gonadotropin stimulation in mice.
77
Technologies). Quantitative polymerase chain reaction (PCR) was performed using a light cycler 480 (Roche, Nutley, NJ, USA) with 480 SYBR Green I Master (Roche, Nutley, NJ, USA) and the following PCR
Materials and methods Experimental Animal studies were undertaken using four-week-old prepubertal female mice. C57Bl/6J mice were purchased from Harlan Laboratories (Jerusalem, Israel). The murine OHSS model was performed as described previously [21]. Briefly, pregnant mare serum gonadotropin (Sigma–Aldrich) was administered intraperitoneally as a high dose (20 U PMSG for 2 days), low dose (5 U PMSG for 1 day) or sham (control) injection (0 U PMSG). hCG 5 U was injected on Day 3 to induce ovulation. Mice were killed 48 h after hCG administration, and the ovaries were harvested. All animal procedures were performed in compliance with the guidelines of Weizmann Institute of Science, and protocols were approved by the Institutional Animal Care and Use Committee. Flow cytometry Ovarian specimens were digested with an enzyme mixture including DNase-I 100 mg/ml (Roche, Mannheim, Germany) and collagenase type IV 1 mg/ml (Sigma–Aldrich, Israel), dissolved in phosphate-buffered saline (PBS), at 37 C for 30 min. Digested tissue was filtered through a 40 mM cell strainer and resuspended in fluorescence-activated cell-sorting buffer (PBS, EDTA5 mM and 3% fetal calf serum). Immunostaining was performed in the presence of rat anti-mouse Fc-gamma receptor III/II (CD16/32; eBioscience, San Diego, CA, USA) by incubating the cells with monoclonal antibodies for 30 min on ice. Staining reagents included fluorochrome (APC, PE or FITC)-labelled anti-CD11b, CD45, Gr1, CD11c and MHC-II (all purchased from eBioscience). Flow cytometry was performed with a FACSORT (Becton Dickinson, Mountain View, CA, USA) and analysed using Flowjo Software (Tree Star Inc., Ashland, OR, USA). IMCs were defined as CD45+CD11b+Gr1+ cells and dendritic cells were defined as CD45 + CD11c+MHCIIhigh cells. Immunohistochemistry Frozen sections (12 mm) of ovarian tissue were fixed with acetone ( 20 C), washed in PBS and blocked with 20% horse serum in PBS. Immunostaining was performed with rat antimouse PECAM1 PE (BD Biosciences Pharmingen), biotin antimouse CD11b and biotin anti-mouse CD11c (BD Biosciences Pharmingen). Streptavidin-cy2 (Jackson ImmunoResearch, Suffolk, UK) was used as the secondary antibody for the CD11b and CD11c staining. Sections were mounted on glass slides with ProLong Gold anti-fade reagent and DAPI mounting media (Invitrogen Life Technologies, Burlington, Ontario, Canada). Ovarian morphology and the presence of corpura lutea were determined in sections stained with haematoxylin and eosin. RNA processing and real-time polymerase chain reaction RNA was isolated by EZ-RNA (Biological Industries, Beit Haemek, Israel), according to the manufacturer's instructions. CDNA was amplified with SuperScript II reverse transcriptase (Invitrogen Life
Fig. 2. Recruitment of bone-marrow-derived CD45+ cells into stimulated ovaries. (A) Representative flow cytometry histograms of anti-CD45 stained single cell suspensions, derived from control ovaries (top graph, cont, n = 3), low-dose stimulated ovaries (middle graph, low, n = 5) and high-dose stimulated ovaries (bottom graph, high, n = 4) (FSC: forward scatter). (B) Bar histogram depicts the percentages of CD45+ cells among the ovarian cells described in (A). Data represent mean standard deviation. *p < 0.05 (Mann–Whitney test).
78
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
primers: S100a8, F-AAATCACCATGCCCTCTACAAG, R-CCCACTTTTATCACCATCGCAA; S100a9, F-ATACTCTAGGAAGGAAGGACACC, R-TCCATGATGTCATTTATGAGGGC; Dmbt1, F-GAGGACTATCTGTGGAGATG, R-TTTGTCCCTCCTGGATTCCA; Vcan, F-TGGGATTGAAGACACTCAGG, R-ATCAGACAGCCATCCTGCAT; Mmp12, F-CTGCTCCCATGAATGACAGTG, R-AGTTGCTTCTAGCCCAAAGAAC; and Actb (used for normalizing control), F-GGCTGTATTCCCCTCCATCG, R-CCAGTTGGTAACAATGCCATGT.
Statistical analysis Data presented in the histograms are expressed as mean standard deviation. Significance of the differences between continuous variables was determined using the two-tailed Mann–Whitney non-parametric test. Real-time PCR expression data were analysed using the unpaired two-tailed Student's t-test. p < 0.05 was considered to indicate significance.
Fig. 3. Ovarian stimulation results in enrichment of immature myeloid cells (IMCs) and depletion of dendritic cells (DCs). Single cell suspensions, derived from control (n = 3), low-dose stimulated ovaries (n = 5) and high-dose stimulated ovaries (n = 4) were analysed by flow cytometry. (A) Representative images of the proportion of CD11b+Gr1+ IMCs among the CD45+ cells (gating as in Fig. 2A) in control ovaries (top graph), low-dose hyperstimulated ovaries (middle graph) and high-dose hyperstimulated ovaries (bottom graph). Bar histogram depicts the percentages of CD11b+Gr1+ IMCs within ovaries. (B) Representative images of the proportion of CD11c+MHCIIhigh dendritic cells among the CD45+ cells in control ovaries (top graph), low-dose hyperstimulated ovaries (middle graph) and high-dose hyperstimulated ovaries (bottom graph). Bar histogram depicts the percentages of CD11c+MHCII+ dendritic cells within ovaries. Data represent mean standard deviation. *p < 0.05 (Mann–Whitney test).
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Results Ovarian stimulation leads to a dose-dependent enrichment of ovarian IMCs paralleled by a reciprocal reduction of ovarian dendritic cells The effect of gonadotropin stimulation on gross weight, ovarian weight and histological appearance was examined. Mice in both stimulated groups demonstrated a slight, yet significant, increase in weight: 14.2 0.8 g (control), 16.1 1.3 g (low dose) and 15.2 0.9 g (high dose) (p < 0.05). Ovaries harvested 48 h after hCG treatment in mice that were treated with low-dose (5 U PMSG) and high-dose (20 U PMSG) gonadotropins were significantly larger and demonstrated more than a four-fold increase in weight compared with sham-treated control ovaries (Fig. 1A, C): 4.4 2.4 mg (control), 22.1 0.8 mg (low dose) and 21.7 5.8 g (high dose) (p < 0.05). Multiple large follicles and numerous postovulatory corpora lutea were identified (Fig. 1B). No significant differences in weight and general appearance of the ovaries were noticed between low- and high-dose groups (Fig. 1A, C). Analysis of ovarian cells by flow cytometry demonstrated a small, yet significant, increase in the ovarian CD45+ bonemarrow-derived cell population in both low- and high-dose groups compared with the control group (Fig. 2): 1.6 0.7% (control), 4.5 1.2% (low dose) and 3.0 1.7% (high dose) (p < 0.05). Importantly, upon ovarian stimulation, the proportion of CD11b+Gr1+ IMCs among the CD45+ bone-marrow-derived cells in the ovaries was higher compared with the control ovaries. Of note, this increase in the ovarian CD11b+Gr1+ IMC population was
79
dose-dependent, as approximately a two-fold increase and a fivefold increase were observed following low- and high-dose gonadotropin treatment, respectively (Fig. 3A): 4.1 0.3% (control), 7.2 1.6% (low dose) and 22.6 8.1% (high dose) (p < 0.05). Concomitantly, a dose-dependent reduction in the proportion of CD11c+MHCII+ dendritic cells among the CD45+ bone-marrowderived cells was noted, as approximately a two-fold and a threefold decrease in this cell population was demonstrated following low- and high-dose gonadotropin treatment, respectively (Fig. 3B): 27.3 8.2% (control), 20.7 4.8% (low dose) and 15.1 1.9% (high dose) (p < 0.05). This finding was also confirmed by immunohistochemistry, with a marked increase in IMCs paralleled by a decrease in dendritic cells upon ovarian stimulation (Fig. 4A, B). Of note, upon ovarian stimulation, IMCs seem to be recruited to an area adjacent to the follicular wall (Fig. 4A), whereas in unstimulated ovaries, dendritic cells are localized sparsely within the stroma (Fig. 4B). Ovarian stimulation is associated with differential expression of key angiogenic genes The global expression pattern of IMCs has been characterized previously in both normal and malignant processes in which these cells are required to support angiogenesis [11,22,23]. This study examined the expression of key genes that have previously been shown to play roles in angiogenesis, and were regulated differentially in IMCs from dormant vs angiogenic tumours, or tumours vs placentas [24–31]. The results indicated that the
Fig. 4. Immature myeloid cells (IMCs) are recruited to the follicular wall upon ovarian stimulation with pregnant mare serum gonadotropins (PMSG). Frozen sections of control and high-dose stimulated ovarian tissues were immunostained with anti-PECAM1 antibody (orange), anti-CD11b (IMCs, green) (A) and anti-CD11c (dendritic cells, green) (B). 4'-6-Diamidino-2-phenylindole (DAPI) (blue) was used to detect nuclei.
80
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
Fig. 5. Differential gene expression in control (WT: wild type) vs high-dose stimulated ovaries (ovarian hyperstimulation syndrome, OHSS). Quantitative reverse transcriptase polymerase chain reaction analysis of relative expression of the indicated genes in control vs high-dose hyperstimulated ovaries. Data represent mean standard deviation of experiments performed in triplicate. *The differences in upregulated genes presented in (A) and downregulated genes presented in (B) were significant (Student's t-test, p < 0.05). Genes presented in (C) did not demonstrate a significant change in expression between the control and high-dose hyperstimulated ovaries.
expression of S100a8, S100a9, Dmbt1 and Vcan was significantly upregulated in stimulated ovaries compared with control ovaries (Fig. 5A). On the other hand, the expression of Mmp12 was significantly downregulated in stimulated ovaries compared with control ovaries (Fig. 5B). Of note, the expression of other genes that are differentially expressed in different IMC populations and known to play key roles in angiogenesis, including Il1b, Jag1, Twist1, Tgfbr3, Klk1, Serpine1 and Ppbp, did not change significantly upon ovarian hyperstimulation (Fig. 5C). Comment The central feature of OHSS is the development of vascular hyperpermeability. As hCG causes the ovary to undergo extensive luteinization, large amounts of oestradiol and progesterone are released, along with other local cytokines. Various vasoactive substances such as VEGF, tumour necrosis factor-a, endothelin-1 and others are secreted by the multiple corpora lutea [3,32]. These, in turn, lead to vascular hyperpermeability by making local
capillaries ‘leaky’, leading to a shift of fluids from the intravascular system to the abdominal and pleural cavities. Nevertheless, the cellular mechanisms that govern this complex process (i.e. which specific cells within the ovary release these vasoactive substances) is still largely unknown. This study used a murine model to examine the alterations in myeloid cell populations within the ovaries upon ovarian stimulation. Both low- and high-dose stimulation were used to control for dose-dependent effects. Both low- and high-dose stimulation led to a similar marked increase in ovarian size and weight, and the appearance of multiple mature follicles and corpora lutea. Of note, the increase in total body weight of the mice was small. It is possible that a more significant increase in weight would have been demonstrated if higher doses of gonadotropins were administered. However, this was not done in the present study in order to reduce unnecessary animal suffering. Upon ovarian stimulation, a small, yet significant, increase in the percentage of CD45+ bone-marrow-derived cells in the ovaries
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
was noted. These cells represent the total haematopoietic cell compartment within the ovaries, and include different cell populations including IMCs and dendritic cells. This mild increase may represent a net influx of blood cells into the ovaries during ovarian stimulation. Notably, this increase was observed in mice stimulated with both low- and high-dose gonadotropins. Although there was a slight trend for an increase in the low-dose group compared with the high-dose group, this trend was insignificant and likely to be of no biological significance. The striking dose-dependent increase in ovarian CD11b+Gr1+ IMCs in parallel to the decrease in CD11c+MHC-II+ dendritic cells upon stimulation suggests an intriguing notion that a shift in myeloid populations at different levels of maturation may affect ovarian angiogenesis, blood vessel permeability and, as a result, the severity of OHSS. IMCs populate the unstimulated ovary in very low numbers, while gonadotropin stimulation leads to a more than five-fold increase in this cell population. This signifies the importance of these cells in this process, presumably as regulators of angiogenesis and increased vascular permeability. The fact that IMCs localize to the vicinity of the follicular wall, a site known for intense angiogenesis within the ovary, suggests that IMCs in the ovary also play an active role in angiogenesis, as described in pregnancy [10,11] and in cancer [18,23]. Additionally, the differential location of dendritic cells (sparsely within the stroma) and IMCs (follicular wall) further implies that these are different cells with different physiological roles within the ovary. Previous studies [18,19,33,34] have shown that angiogenesis in general (as demonstrated in Matrigel plugs in vivo) and angiogenesis observed in tumours, endometriosis and choroidal neovascularization is dependent on the presence of immature dendritic cells. Specifically, supplementation of immature dendritic cells, but not mature dendritic cells, enhanced lesion growth. Depletion of dendritic cells in a transgenic mouse model that allows for their conditional ablation completely abrogated angiogenesis in Matrigel plugs, decreased tumour growth and decreased the growth of endometriotic lesions. Accordingly, the present study demonstrated that increased angiogenesis in the ovary is associated with ovarian stimulation, and ovulation is accompanied by a decrease in non-angiogenic mature dendritic cells. Nevertheless, the presence of proangiogenic immature dendritic cells in the ovary was not detected. Interestingly, the authors have recently demonstrated [35] that labour and delivery are preceded by similar myeloid cell alterations, reflected by a decrease in IMCs and an increase in dendritic cells populating the mouse placenta. It is therefore tempting to speculate that a governing mechanism applies both in midtrimester placentas and stimulated ovaries, involving the presence of IMCs and immature dendritic cells when active angiogenesis takes place. It is possible that when angiogenesis is no longer needed, IMCs mature into more differentiated cells such as dendritic cells. Increased expression of key genes, shown to be highly expressed by IMCs and to play a role in angiogenesis, further supports the notion that these cells contribute to vascular permeability in the pathogenesis of OHSS. S100a8 and S100a9, known to be involved in angiogenesis [36], were both highly expressed in placental- and tumour-derived IMCs [23]. Dmbt1 and Vcan, also key players in angiogenesis [25,26], were highly expressed in tumour-derived IMCs [22]. Although the global gene expression pattern of ovarian IMCs has not been described to date, the fact that Mmp12 is highly expressed by IMCs in other systems and downregulated upon ovarian stimulation, and that other key IMC genes such as Il1b, Jag1, Twist1, Tgfbr3, Klk1, Serpine1 and Ppbp were unaffected suggests that the ovarian IMC gene expression pattern is unique. Overall, these data suggest that changes in proangiogenic myeloid populations within the ovary may play a role in ovarian
81
hyperstimulation. Key angiogenic factors are expressed within recruited IMCs, and may influence vascular permeability and angiogenesis. Studies testing therapies that target these different myeloid cell populations as a novel strategy for the prevention and treatment of OHSS in humans are warranted. References [1] Avecillas JF, Falcone T, Arroliga AC. Ovarian hyperstimulation syndrome. Crit Care Clin 2004;20:679–95 ix. [2] Rydberg E, Pedersen-Bjergaard K. Effect of serum gonadotropin and chorionic gonadotropin on the human ovary. JAMA 1943;121:1117–22. [3] Grossman LC, Michalakis KG, Browne H, Payson MD, Segars JH. The pathophysiology of ovarian hyperstimulation syndrome: an unrecognized compartment syndrome. Fertil Steril 2010;94:1392–8. [4] Levin ER, Rosen GF, Cassidenti DL, et al. Role of vascular endothelial cell growth factor in ovarian hyperstimulation syndrome. J Clin Invest 1998;102:1978–85. [5] Navot D, Margalioth EJ, Laufer N, et al. Direct correlation between plasma renin activity and severity of the ovarian hyperstimulation syndrome. Fertil Steril 1987;48:57–61. [6] Garcia-Velasco JA, Pellicer A. New concepts in the understanding of the ovarian hyperstimulation syndrome. Curr Opin Obstet Gynecol 2003;15:251–6. [7] Scotti L, Abramovich D, Pascuali N, et al. Involvement of the ANGPTs/Tie-2 system in ovarian hyperstimulation syndrome (OHSS). Mol Cell Endocrinol 2012;365:223–30. [8] Soares SR. Etiology of OHSS and use of dopamine agonists. Fertil Steril 2012;97:517–22. [9] Budev MM, Arroliga AC, Falcone T. Ovarian hyperstimulation syndrome. Crit Care Med 2005;33(Suppl. 10):S301–6. [10] Mei-Dan E, Fainaru O, Hantisteanu S, Shavit T, Ellenbogen A, Hallak M. Proangiogenic immature myeloid cells populate the human placenta and their presence correlates with placental and birthweight. Am J Obstet Gynecol 2012;207:141e1–5. [11] Fainaru O, Hantisteanu S, Hallak M. Immature myeloid cells accumulate in mouse placenta and promote angiogenesis. Am J Obstet Gynecol 2011;204:544e18–23. [12] Yang L, DeBusk LM, Fukuda K, et al. Expansion of myeloid immune suppressor Gr+CD11b+ cells in tumor-bearing host directly promotes tumor angiogenesis. Cancer Cell 2004;6:409–21. [13] Yan HH, Pickup M, Pang Y, et al. Gr1+CD11b+ myeloid cells tip the balance of immune protection to tumor promotion in the premetastatic lung. Cancer Res 2010;70:6139–49. [14] Shojaei F, Wu X, Malik AK, et al. Tumor refractoriness to anti-VEGF treatment is mediated by CD11b+Gr1+ myeloid cells. Nat Biotechnol 2007;25:911–20. [15] Banchereau J, Steinman RM. Dendritic cells and the control of immunity. Nature 1998;392:245–52. [16] Sozzani S, Rusnati M, Riboldi E, Mitola S, Presta M. Dendritic cell–endothelial cell cross-talk in angiogenesis. Trends Immunol 2007;28:385–92. [17] Murdoch C, Muthana M, Coffelt SB, Lewis CE. The role of myeloid cells in the promotion of tumour angiogenesis. Nat Rev Cancer 2008;8:618–31. [18] Fainaru O, Almog N, Yung CW, et al. Tumor growth and angiogenesis are dependent on the presence of immature dendritic cells. FASEB J 2010;24:1411–8. [19] Fainaru O, Adini A, Benny O, et al. Dendritic cells support angiogenesis and promote lesion growth in a murine model of endometriosis. FASEB J 2008;22:522–9. [20] Satpathy AT, Wu X, Albring JC, Murphy KM. Re(de)fining the dendritic cell lineage. Nat Immunol 2012;13:1145–54. [21] Fainaru O, Hornstein MD, Folkman J. Doxycycline inhibits vascular leakage and prevents ovarian hyperstimulation syndrome in a murine model. Fertil Steril 2009;92:1701–5. [22] Pencovich N, Hantisteanu S, Wurtzel O, Hallak M, Fainaru O. Unique expression patterns associated with preferential recruitment of immature myeloid cells into angiogenic versus dormant tumors. Genes Immun 2012;14:90–8. [23] Fainaru O, Pencovich N, Hantisteanu S, Yona G, Hallak M. Immature myeloid cells derived from mouse placentas and malignant tumors demonstrate similar proangiogenic transcriptional signatures. Fertil Steril 2013;99:910–7. [24] Chen P, Huang Y, Bong R, et al. Tumor-associated macrophages promote angiogenesis and melanoma growth via adrenomedullin in a paracrine and autocrine manner. Clin Cancer Res 2011;17:7230–9. [25] Muller H, Hu J, Popp R, et al. Deleted in malignant brain tumors 1 is present in the vascular extracellular matrix and promotes angiogenesis. Arterioscler Thromb Vasc Biol 2012;32:442–8. [26] Vazquez-Ortiz G, Pina-Sanchez P, Vazquez K, et al. Overexpression of cathepsin F, matrix metalloproteinases 11 and 12 in cervical cancer. BMC Cancer 2005;5:68. [27] Min JK, Park H, Choi HJ, et al. The WNT antagonist dickkopf2 promotes angiogenesis in rodent and human endothelial cells. J Clin Invest 2011;121:1882–93. [28] Cheng P, Jiang FH, Zhao LM, et al. Human macrophage metalloelastase correlates with angiogenesis and prognosis of gastric carcinoma. Dig Dis Sci 2010;55:3138–46. [29] Hu L, Roth JM, Brooks P, Ibrahim S, Karpatkin S. Twist is required for thrombininduced tumor angiogenesis and growth. Cancer Res 2008;68:4296–302.
82
N. Pencovich et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 179 (2014) 75–82
[30] Sahin H, Borkham-Kamphorst E, Kuppe C, et al. Chemokine Cxcl9 attenuates liver fibrosis-associated angiogenesis in mice. Hepatology 2012;55:1610–9. [31] Dong M, How T, Kirkbride KC, et al. The type III TGF-beta receptor suppresses breast cancer progression. J Clin Invest 2007;117:206–17. [32] Kumar P, Sait SF, Sharma A, Kumar M. Ovarian hyperstimulation syndrome. J Hum Reprod Sci 2011;4:70–5. [33] Pencovich N, Luk J, Hantisteanu S, Hornstein MD, Fainaru O. The development of endometriosis in a murine model is dependent on the presence of dendritic cells. Reprod Biomed Online 2014;28:515–21.
[34] Nakai K, Fainaru O, Bazinet L, et al. Dendritic cells augment choroidal neovascularization. Investig Ophthalmol Vis Sci 2008;49:3666–70. [35] Fainaru O, Paz G, Hantisteanu S, Hallak M. Myeloid cell alterations in the mouse placenta precede the onset of labor and delivery. Am J Obstet Gynecol 2014;210:73e1–7. [36] Li C, Li S, Jia C, Yang L, Song Z, Wang Y. Low concentration of S100A8/9 promotes angiogenesis-related activity of vascular endothelial cells: bridges among inflammation, angiogenesis, and tumorigenesis? Mediators Inflamm 2012;2012:248574.