Skip to main content
NCBI home page
Endocrinology logoLink to Endocrinology
. 2015 Aug 17;156(11):4293–4301. doi: 10.1210/en.2015-1522

Protective Hematopoietic Effect of Estrogens in a Mouse Model of Thrombosis: Respective Roles of Nuclear Versus Membrane Estrogen Receptor α

Marie-Cécile Valéra 1,✉, Coralie Fontaine 1, Françoise Lenfant 1, Cendrine Cabou 1, Maeva Guillaume 1, Natalia Smirnova 1, Sung Hoon Kim 1, Pierre Chambon 1, John A Katzenellenbogen 1, Benita S Katzenellenbogen 1, Bernard Payrastre 1,*, Jean-François Arnal 1,*,✉
PMCID: PMC4606746  PMID: 26280130

Abstract

We recently reported that chronic 17β-estradiol (E2) treatment in mice decreases platelet responsiveness, prolongs the tail-bleeding time and protects against acute thromboembolism via the hematopoietic estrogen receptor alpha (ERα), and independently of ERβ. Here, we have explored the respective roles of membrane vs nuclear actions of ERα in this process, using: 1) the selective activator of membrane ERα: estrogen dendrimer conjugate, and 2) mouse models with mutations in ERα. The selective targeting of activation function 2 of ERα provides a model of nuclear ERα loss-of-function, whereas mutation of the ERα palmitoylation site leads to a model of membrane ERα deficiency. The combination of pharmacological and genetic approaches including hematopoietic chimera mice demonstrated that absence of either membrane or nuclear ERα activation in bone marrow does not prevent the prolongation of the tail-bleeding time, suggesting a redundancy of these two functions for this E2 effect. In addition, although hematopoietic membrane ERα is neither sufficient nor necessary to protect E2-treated mice from collagen/epinephrine-induced thromboembolism, the protection against death-induced thromboembolism is significantly reduced in the absence of hematopoietic nuclear ERα activation. Overall, this study emphasizes that hematopoietic cells (likely megakaryocytes and possibly immune cells) constitute an important target in the antithrombotic effects of estrogens, and delineate for the first time in vivo the respective roles of membrane vs nuclear ERα effects, with a prominent role of the latter.

Estrogens play a pivotal role in reproduction and during the menstrual cycle, and they modulate cancer growth, bone demineralization and vascular risks (1–4). Although the effect of estrogens on coagulation and venous thrombosis risk has received much attention (5), their effect on platelet function remains poorly characterized. Platelets are small anuclear cell fragments derived from bone marrow megakaryocytes that play a critical role in hemostasis and thrombosis. Hemostasis is the physiological process of formation of a platelet clot to arrest bleeding following blood vessel injury, whereas thrombosis is the pathological formation of a blood clot obstructing the vessel and involving both blood cells and clotting factors (6). Platelets also play a role in the maintenance of the integrity of the endothelium, in part through the release of proangiogenic cytokines and growth factors, as well as through their effect on immune cells (7).

The biological effects of 17β-estradiol (E2), the main endogenous estrogen, are mediated by its binding to two nuclear steroid receptors: estrogen receptor alpha (ERα) and beta (ERβ), leading to conformational changes and in turn to the recruitment of coactivators and subsequent interaction to estrogen response elements (8). Ligand-induced transcriptional activity of ER involves the action of two distinct activation functions (AF), AF1 and AF2, located in the N-terminal A/B and C-terminal E domains, respectively. The time lag between estrogen administration and observable transcriptional effects is typically in the order of hours to days. However, in addition to the genomic actions of ER, E2 has been found to induce rapid effects occurring within minutes following administration. These effects are mediated through a subpopulation of receptors associated with the plasma membrane, a process usually termed “membrane-initiated steroid signaling,” “nongenomic,” or “extranuclear” effects (9), and they lead to several cell signaling responses, such as mobilization of intracellular calcium and modulation of various protein kinase pathways (10).

To exclude the involvement of nuclear ER, membrane-impermeable conjugates of E2, such as E2-BSA, have been widely used in the past (11–13). These cell-impenetrant macromolecules were supposed to interact only with ERα associated with the plasma membrane. However, there are reports suggesting that E2-BSA is not an well suited for investigating membrane estrogen receptors (12, 14): 1) The site selected to link the estrogen to the BSA can interfere with hormone binding to ER (12), and 2) solutions of E2-BSA are not stable and contain free E2 or small fragments of BSA that result from degradation of the protein carrier by the cells (14). In this context, a new class of estrogen-macromolecule conjugates, termed estrogen dendrimer conjugates (EDCs), was developed (15). EDC consists of ethinyl estradiol attached to a large, abiological, and positively charged nondegradable poly(amido)amine dendrimer via hydrolytically stable linkages (15). EDC affords optimal ligand access to ER and has a binding affinity comparable to that of the native ligand (16, 17). This new conjugate has a high chemical stability, is free from traces of unattached ligand and is very specific in stimulating nongenomic responses (14, 15). In addition, in vivo administration of EDC promotes endothelial protection, but neither uterine nor breast cancer growth in mice (18), indicating that extranuclear ERα signaling is sufficient to promote the beneficial vascular effects of estrogens.

To investigate the physiological role of membrane-initiated ERα actions in vivo, we recently created a knock-in mouse model by mutating the cysteine 451 palmitoylation site of ERα to alanine (designated C451A-ERα); this mutant mouse can be considered as a model of specific membrane ERα loss-of-function (19). We also developed a complementary mutant mouse model of inactivation of the activation function AF2, consisting in a deletion of the amino acids 543–549 in the helix 12 of ERα (designated ERα-AF20); this mutant mouse provides a selective loss-of-function of nuclear ERα actions, but retains at the same time the membrane ERα function and thereby the endothelial responses to E2 (20, 21).

We previously demonstrated that a chronic high physiologic dose of E2 has an inhibitory effect on platelet responsiveness ex vivo (22). In vivo, E2 treatment increased mice tail-bleeding times and induced resistance to collagen epinephrine–induced thromboembolism, through the hematopoietic ERα and independently of ERβ (22). In the present work, we sought to determine the respective roles of hematopoietic membrane and nuclear ERα in the effects of estrogens on hemostasis and thrombosis. Our results show that, although hematopoietic nuclear ERα has a prominent role, the effect of E2 on the tail-bleeding times and on collagen/epinephrine-induced thromboembolism are mediated by both hematopoietic nuclear and membrane ERα.

Materials and Methods

Experimental animals

Female C57BL/6J mice were purchased from Charles River. Mice deleted for ERα-AF2 and C451A-ERα mice have been described previously (19, 21). All procedures were performed in accordance with the guidelines established by the National Institute of Medical Research. Mice were anesthetized by ip injection of ketamine (25 mg/kg) and xylazine (10 mg/kg) and ovariectomized at 4 weeks of age. Two weeks after ovariectomy, mice were treated with E2 (200 μg∙kg−1·d−1 in 60-d-release pellets), EDC (estradiol dendrimer conjugated, 240 μg∙kg−1·d−1), or empty dendrimer at a rate identical to that delivered with EDC as a control. This high-physiological dose of E2 induces E2 plasma level in the nanomolar range (23), ie, in the high range of pregnant mice, given that adult mice, at variance with women, do not express SHBG (24). The dose of EDC was previously demonstrated to elicit circulating concentrations in the same range (18). The mice were euthanized after a 3-week treatment period.

Bone marrow transplantation

Two weeks after ovariectomy, recipient mice were lethally irradiated (9.2 Gy, γ source) then iv reconstituted with bone marrow cells from either mice with a point mutation of the ERα palmitoylation site (designated C451A-ERα or wild-type (WT)-C451-ERα) or mice mutated for the activation function 2 of ERα (designated ERα-AF20 or ERα-AF2+/+). Three weeks later, transplanted mice were implanted or not with an E2 pellet (200 μg∙kg−1·d−1 in 60-d-release pellets). Sulfamethoxazole (400 mg/mL) and trimethoprim (80 mg/mL) was added to their drinking water for 3 weeks after bone marrow transplantation.

Materials

Collagen Reagent HORM (type I fibrils from equine tendon) suspension was purchased from Takeda, DIOC from Life Technologie, fibrinogen from Sigma-Aldrich and FeCl3 from Mallinckrodt Chemical. The EDC, derivatized with ethinyl estradiol in a 20:1 ligand/dendrimer stoichiometry, and its control molecule, the empty dendrimer were synthesized as previously described (15, 16), stored in methanol at −20°C, and used within 3 months of preparation.

Tail-bleeding time

After mice anesthesia, a 3-mm segment of tail tip was cut off and the bleeding time was monitored as described previously (25). Experiments were terminated after 30 minutes if no cessation of blood flow occurred.

Thromboembolism

Acute systemic vascular thromboembolism was induced by injecting a mixture of collagen (0.4 mg/kg) and epinephrine (60 μg/kg) into the right jugular vein of anesthetized mice as described (25).

Histochemical analysis of mouse lung

To visualize thrombi in the pulmonary vasculature, anesthetized mice were euthanized 10 minutes after injection of the collagen-epinephrine mixture. Lungs were excised and formalin fixed. Paraffin sections (5-μm thick) were stained with hematoxylin-eosin and analyzed. Platelets were identified with a rabbit anti-αIIb integrin Ab.

Carotid artery thrombosis

The right and left carotids were dissected free from surrounding tissues. Two flow probes were connected to a Transonic model T403 flow meter (Transonic System; Emka Technologies) to record the blood flow (milliliters per minute) of the carotids. FeCl3 was used to induce vascular injury. A 1 × 4-mm strip of paper saturated with 7% FeCl3 solution was applied to the adventitial surface of the left carotid for 2 minutes then removed. The right carotid was used as an internal control. Blood flow was then monitored continuously throughout the procedure (IOX software) (22).

Ex vivo flow-based adhesion studies

Biochips with microcapillaries (Vena8Fluoro+, Cellix) were coated with type I collagen from equine tendon (500 μg/mL) or fibrinogen (100 μg/mL) for 1 hour at 37°C. Blood was drawn into heparin (10 IU/mL), and DiOC6 (2μM, 30 min at 37°C) was used to label platelets in whole blood. Labeled blood was then perfused through microcapillaries using a syringe pump (PHD-2000; Harvard Apparatus) at the indicated wall shear rate (20, 60, or 160 dynes/cm2) for 2 minutes. Platelet adhesion was visualized with a 40× oil immersion objective in real time (acquisition rate: 1 frame every 5 sec) for both fluorescent and transmitted light microscopy. Analysis of surface coverage was performed offline on single frame by quantification of pixel surface by thresholding using ImageJ software.

Statistics

Results are expressed as mean ± SD or SEM, as indicated. Statistical analyses were performed using GraphPad (Student t test). P values < .05 were considered statistically significant.

Results

EDC treatment prolongs the tail bleeding time without affecting the coagulation tests and the platelet count

We first analyzed by a pharmacological approach the contribution of membrane ERα on primary hemostasis and thrombosis using the selective agonist of the membrane ERα, EDC. Primary hemostasis, in which platelets play a key role, was investigated in vivo by measuring tail-bleeding times. In untreated ovariectomized mice, the bleeding time was in a normal range (5.2 min; SD ± 2.4; n = 12) (Figure 1A). Consistent with our previous data, E2 treatment prolonged the tail-bleeding time, which varied from a few minutes to times in excess of 30 minutes (22). EDC treatment also increased the tail-bleeding time with some heterogeneity. Indeed, among 22 mice tested, bleeding was weakly prolonged (> 10 min) in nine cases and increased (> 15 min) in four cases. EDC was, however, less efficient than E2 in inducing prolonged bleeding time given that 18.2% of mice bled greater than 15 minutes following EDC treatment vs 60% upon E2. EDC treatment had no effect on coagulation tests (prothrombin time and activated partial thromboplastin time) or on the levels of plasma fibrinogen and coagulation factors (Supplemental Table 1A). Furthermore, platelet as well as white and red blood cells counts were not affected by the treatment (Supplemental Table 1B). Thus, the effect of EDC on primary hemostasis was not due to functional deficiencies in coagulation or modification of platelet counts.

Figure 1.

Figure 1.

Open in a new tab

Increased bleeding time in EDC-treated mice. Tail bleeding times of ovariectomized (OVX) (n = 12), EDC-treated mice (+EDC) (n = 22) and E2-treated mice (+E2) (n = 15) were monitored. The experiment was stopped after 30 minutes if no cessation of blood flow occurred. Each point represents one individual.

Effect of EDC treatment on thrombus formation under flow ex vivo

Upon vascular injury, platelets adhere to the exposed subendothelial matrix via specific adhesive glycoproteins. To mimic the interaction between platelets and subendothelial matrix, we used an ex vivo flow-based adhesion assay where fluorescently labeled platelets in whole blood were perfused over two common extracellular matrix proteins, collagen and fibrinogen. Adhesion of platelets to collagen leads to thrombus formation whereas platelet adhesion to fibrinogen results in a monolayer of platelets, mainly due to αIIbβ3-fibrinogen interaction. The use of microcapillary to test platelet adhesion and thrombus formation under shear flow in whole blood is a well-established technique for evaluating platelet functions in an integrated manner (26, 27).

We first analyzed thrombus formation on collagen ex vivo under arterial flow conditions (20 and 60 dynes/cm2). Under high arterial shear rate (60 dynes/cm2), corresponding to the wall shear rate found in arterioles, the surface occupied by adhered and aggregated platelets was similar in control and EDC-treated mice (Figure 2A). Conversely, under these conditions, E2 treatment reduced platelet adhesion and thrombus growth on collagen, as previously described (22). At low arterial shear rate (20 dynes/cm2), a significant decrease in the percentages of surface covered by platelets was observed both in EDC and E2-treated mice compared with control (Figure 2B). To assess the stability of the thrombi formed, the shear rate was rapidly increased from 20 to 160 dynes/cm2 during 2 minutes. Under these conditions, platelet aggregates were stable and no platelet embolization could be observed (Figure 2B). However, at high shear rate (160 dynes/cm2), the slope of the curve tended to be less in platelets from ovariectomized mice in comparison with platelets from EDC and E2-treated mice, suggesting that the absence of hormones could effect platelet adhesion at this shear rate.

Figure 2.

Figure 2.

Open in a new tab

Thrombus formation under arterial flow conditions ex vivo. DiOC6-labeled platelets in whole blood were perfused over a collagen-coated microcapillary at a wall shear rate of (A) 60 dynes/cm2 for 2 min, or (B) 20 dynes/cm2 for 2 min. Thrombus formation was visualized with a 40× long working distance objective in real time and then imaged by transmitted light microscopy. Area covered by platelet thrombi were measured at 2 surface locations in each of three or four different experiments (mean ± SEM). Representative fluorescent images taken at the end of the experiment are shown. C, DIOC6-labeled platelets in whole blood were perfused through fibrinogen-coated plates at a shear rate of 20 dynes/cm2 for 2 min. Area covered by platelets was measured (mean ± SEM of four experiments). Representative fluorescent images taken at the end of the experiment are shown.

When platelet adhesion on fibrinogen matrix was analyzed, control, EDC, and E2-treated platelets formed a comparable regular platelet monolayer (Figure 2C), stable even when a shear rate of 160 dynes/cm2 was applied for 2 minutes (data not shown). Overall, these data suggest that EDC has no effect on platelet adhesion on fibrinogen but decreases platelet thrombus formation on collagen fibers under low arterial shear rate. EDC is, however, less efficient than E2 and has no effect on thrombus formation at high shear rate.

Effect of EDC treatment on two models of thrombosis in vivo

To evaluate the effect of chronic EDC treatment in vivo, we used a model of arterial thrombosis. In this model, a carotid artery injury was induced by FeCl3 and the blood flow and time to artery occlusion were determined. All control animals and 6/7 EDC-treated mice (85.7%) presented a complete and stable occlusion within 15 minutes (Figure 3A). By comparison, we previously reported that E2 treatment induced unstable occlusions in 50% of cases and complete resistance to occlusion in 10% of cases (22).

Figure 3.

Figure 3.

Open in a new tab

Thrombus formation in vivo in EDC-treated mice. A, Thrombotic response of mice to ferric chloride injury of the carotid artery. Flow rates were measured in the carotid artery after exposure to 7% FeCl3 for 2 min. Mean ± SD. B, Thromboembolism was induced by injection of a collagen (0.4 mg/kg) and epinephrine (60 μg/kg) mixture into the jugular vein. All EDC-treated mice died within 5 minutes. E2-treated mice (200 μg∙kg−1∙d−1) were protected from thromboembolism. Representative sections of hematoxylin-eosin–stained lungs from a control mouse and an EDC-treated mouse. Original magnification ×100. Platelet staining (αIIb in brown) of lung sections from mice treated or not with EDC. Original magnification, ×100.

In a second model of in vivo acute thrombosis consisting in injection of a combination of a high dose of collagen (0.4 mg/kg) and epinephrine (60 μg/kg) into the jugular vein, EDC-treated mice were not protected against thromboembolism (100% of mortality after 5 min), whereas all E2-treated mice were still alive even 10 minutes after administration (Figure 3B). Histologic examination of lung tissue from control and EDC-treated mice confirmed occlusive pulmonary thrombi throughout the vasculature, particularly in large vessels (Figure 3B). Overall, these in vivo data suggest that activation of membrane ERα is insufficient to mediate the protective effect of estrogen on occlusive thrombus formation.

Increased tail-bleeding times in E2-treated C451A-ERα or ERαAF20 bone marrow chimeras

To further analyze the effect of membrane vs nuclear hematopoietic ERα on primary hemostasis and thrombosis, a genetic approach was used. We generated hematopoietic chimeras with medullar cells from C451A-ERα or ERα-AF20 mice by engrafting lethally irradiated WT ovariectomized mice with bone marrow from the mutant mice (C451A-ERα and ERα-AF20 mice) or their control littermates (WT-C451-ERα and ERα-AF2+/+ mice, respectively). The success of the bone marrow transplantation was confirmed by PCR. Increased uterine weight was observed in E2-treated mice, whereas EDC-treated mice had atrophied uteri (data not shown).

As expected, the tail-bleeding times of most control mice engrafted with wild-type bone marrow were prolonged after E2 treatment. We previously demonstrated that this E2 effect was totally lost in ERα−/− bone marrow chimeras (22). Interestingly, C451A-ERα and ERα-AF20 bone marrow chimeras treated with E2 exhibited a prolonged bleeding time (Figure 4, A and B), indicating that while hematopoietic ERα is absolutely required for this E2 action, either membrane ERα alone or nuclear ERα alone can mediate this effect. However, in C451A-ERα chimeras, 100% of E2-treated mice bled more than 15 minutes vs only 52.9% (nine of 17) in ERα-AF20 mice. Thus, hematopoietic nuclear effects are more important than hematopoietic membrane ERα to get prolonged tail-bleeding time in E2-treated mice.

Figure 4.

Figure 4.

Open in a new tab

Bleeding time in ERα-AF20 and C451A-ERα hematopoietic chimeric mice. Tail-bleeding times of mice (WT) engrafted with bone marrow from (A) WT-C451-ERα or C451A-ERα, or (B) ERα-AF2+/+ or ERα-AF20 and treated or not with E2.

Protection from thromboembolism in E2-treated C451A-ERα or ERαAF20 bone marrow chimeras

We then analyzed the protective effect of E2 against thromboembolism using the same genetic approach. In agreement with previous experiments, the protective effect of E2 against thromboembolism was fully observed in wild-type (WT-C451-ERα and ERα-AF2+/+) bone marrow chimeras. All E2-treated mice reconstituted with C451A-ERα were protected (Figure 5A), demonstrating that hematopoietic membrane ERα activation is dispensable to protect mice against acute thromboembolism. Furthermore, EDC-treated mice engrafted with WT or ERα-AF20 bone marrow died within 5 minutes, demonstrating that membrane ERα activation is not sufficient to mediate the effect of estrogens on thromboembolism (data not shown).

Figure 5.

Figure 5.

Open in a new tab

Thromboembolism in ERα-AF20 and C451A-ERα hematopoietic chimeric mice. Mice engrafted with normal bone marrow or bone marrow from (A) WT-C451-ERα or C451A-ERα or (B) ERα-AF2+/+ or ERα-AF20 mice were treated or not with E2, and thromboembolism assays were performed. Six of 18 mice engrafted with ERα-AF20 bone marrow died (P = .0004 by Fisher's exact test vs ovariectomized mice).

Finally, a majority (12 of 18) of the E2-treated mice reconstituted with ERα-AF20 bone marrow were protected against thromboembolism (Figure 5B), indicating that hematopoietic nuclear ERα is not absolutely necessary for the protective action of estrogens against thromboembolism. However, one third (6 of 18) of these mice with a loss of nuclear ERα function in their hematopoietic cells died from the thromboembolic stress.

Discussion

The present work confirms and further delineates an unexpected target of estrogens that lead to an impressive protection in mouse models of thrombosis (22). Here, we explored the respective roles of membrane vs nuclear actions of ERα in primary hemostasis and thrombus formation, using both pharmacological and genetic approaches. We previously demonstrated that endothelial and uterine actions are dependent on membrane and nuclear ERα, respectively (28, 29). We show here for the first time, that the effect of E2 on the tail-bleeding time and on collagen/epinephrine-induced thromboembolism is mediated by both hematopoietic nuclear and membrane ERα, with a higher effect of the nuclear component.

At variance to E2, treatment of ovariectomized mice with EDC, a selective activator of membrane ERα, did not modify platelet and white blood cells counts and had no effect on the plasma level of fibrinogen and coagulation factors. Similarly, in contrast with E2 treatment that was previously shown to reduce platelet adhesion and thrombus growth on collagen matrix under arterial flow conditions (22), EDC treatment had no effect on platelet adhesion and thrombus growth at high arterial shear rate and a weak effect at low arterial shear rate in ex vivo flow-based adhesion studies. Thus, in comparison with E2 treatment, the effect of EDC is modest, indicating that activation of membrane ERα is not sufficient to mediate the full effect of estrogens on platelet adhesion and thrombus growth under flow.

Whereas E2 markedly protected against collagen/epinephrine-induced thromboembolism (22), we found that mice treated with EDC are neither protected from acute pulmonary thromboembolism nor from occlusive thrombosis after FeCl3 injury of the carotid. These results suggest that the contribution of nuclear ERα is essential for the protective effect of estrogens on thromboembolism. In contrast, both E2- and EDC-treated mice exhibited increased tail-bleeding times, suggesting that nuclear ERα is dispensable for the effect of estrogens on primary hemostasis.

Moreover, two complementary mouse models of selective inactivation of membrane ERα (mutation of ERα palmitoylation site, C451A-ERα) and of nuclear ERα (deletion of seven amino acids in helix 12 leading to AF2 inactivation, ERα-AF20) have been generated in our group these last years (19, 21), in collaboration with the Mouse Clinic in Strasbourg. Using these models, we show that the tail-bleeding time is increased in E2-treated C451A-ERα or ERα-AF20 bone marrow chimeras, demonstrating that both hematopoietic nuclear and membrane ERα are dispensable for this effect of E2 (when the other is present). However, these genetic membrane ERα and nuclear ERα loss-of-function approaches reveal that activation of hematopoietic membrane ERα can partially protect against thromboembolism whereas activation of nuclear ERα confers a complete protection (Table 1).

Table 1.

ERα and Protection Against Thromboembolism

Activation of ERα in: Absence of Protection Partial Protection Total Protection
Hematopoietic cells
    Membrane × ×
    Nuclear ×
Extrahematopoietic cells
    Membrane × × ×
    Nuclear × ×
Open in a new tab

According to our current understanding, the E2 effect in ERα-AF20 mice would be expected to share similarities with EDC effects (ie, activation of only membrane ERα). This expectation was fulfilled in the model of accelerated reendothelization, which probes mainly the in vivo migration of macrovascular endothelial cell monolayer following their injury at the level of the mouse common carotid injury (19). Here, the activation of membrane ERα by EDC was found not to be sufficient to mediate the protective effect of estrogens on thromboembolism, whereas we observed a partial protection in E2-treated mice engrafted with ERα-AF20 bone marrow. This apparent discrepancy could be explained by the effects of E2 on the nuclear ERα of extra hematopoietic cells in mice engrafted with ERα-AF20. However, using ERα−/− chimeras, we have previously demonstrated that the effect of E2 on extrahematopoietic ERα was not sufficient to mediate the effects on platelet responsiveness (22) suggesting a role for both membrane and nuclear extrahematopoietic ERα in this partial protection against thromboembolism.

The comparison between E2 and EDC in term of bioavailability or bioactive concentration is particularly challenging, given that the diffusion, local concentration, and compartmentalization (plasma membrane and subdomains such as caveolae, but also extranuclear actions outside this compartment) of EDC remain to be determined and are certainly different from E2. Indeed, ERα membrane–initiated steroid signaling regulates multiple signaling pathways, in particular through Gαi or striatin interactions (30, 31) and Phosphatidylinositol 3-kinases (32). Thus, EDC might be capable of activating only a subset of all possible extranuclear initiated effects (eg, only those initiated by plasma membrane ERα), whereas E2 could be activating the full range of these extranuclear effects, including plasma membrane and cytoplasmic, this latter one possibly not being accessible to EDC.

Conclusion and perspectives

Our results clearly demonstrated that: 1) selective membrane ERα activation is not sufficient to reduce platelet thrombus formation ex vivo and to mediate the protective effect of estrogens on occlusive thrombus formation in vivo but is sufficient to prolong the tail-bleeding time 2) hematopoietic membrane ERα is not sufficient for a full protection against collagen/epinephrine-induced thromboembolism, and 3) the protective action of E2 on thrombosis is mainly mediated by hematopoietic nuclear ERα effects.

The present work confirms and further delineates an unexpected mechanism of protection against thrombosis by estrogens in these mouse models. It is important to remember that in mice estrogens do not alter the coagulation factors (22) toward a procoagulant profile as they do in women (5). This species difference enabled us in the present study to dissect the molecular mechanisms at the level of ERα subfunctions, and will permit determinations in future studies the respective roles of platelet and white blood cells thanks to Cre-Lox approaches. If we will be able to characterize in humans the hematopoietic counterpart of this major antithrombotic effect of E2 observed in mouse, this could profoundly change our view of the prothrombotic risk of estrogens, which could be determined, along the classic predisposing factors (genetic, venous blood stasis), by the balance between prothrombotic hepatic effects and antithrombotic hematopoietic effects. Finally, these results would help to further delineate what can be expected from new selective estrogen receptor modulators that have been developed or characterized recently as selective membrane or nuclear ERα activators (16).

Acknowledgments

The staff of the animal facilities and of the “Plateforme d'experimentation fonctionnelle” (A. Desquesnes) are acknowledged for their skillful technical assistance. We also thank F. Boudou and M. Buscato for their excellent technical assistance and contribution to qRT-PCR experiments carried out at the GeT-TQ Genopole Toulouse Facility.

The work at INSERM unit U1048 was supported by INSERM, Université de Toulouse III, Faculté de Médecine Toulouse-Rangueil, Fondation de France, Conseil Régional Midi-Pyrénées, and Fondation pour la Recherche Médicale. This work was supported by a National Institutes of Health Grant PHS5R01 DK015556 (to J.A.K.). J-F.A. and B.P are fellows of Institut Universitaire de France.

Disclosure Summary: The authors have nothing to disclose.

Footnotes

Abbreviations:
AF
activation function
C451A-ERα
mutation of the ERα palmitoylation site
E2
17β-estradiol
EDC
estrogen dendrimer conjugate
ERα
estrogen receptor alpha
ERβ
estrogen receptor beta
ERα-AF20
deletion of the activation function 2 of ERα
WT
wild type.

References

  • 1. Manolagas SC, O'Brien CA, Almeida M. The role of estrogen and androgen receptors in bone health and disease. Nat Rev Endocrinol. 2013;9(12):699–712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Brisken C, O'Malley B. Hormone action in the mammary gland. Cold Spring Harb Perspect Biol. 2010;2(12):a003178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Mendelsohn ME. Estrogen actions in the cardiovascular system. Climacteric. 2009;12 Suppl 1:18–21. [DOI] [PubMed] [Google Scholar]
  • 4. Arnal JF, Laurell H, Fontaine C, et al. Estrogen receptor actions on vascular biology and inflammation: Implications in vascular pathophysiology. Climacteric. 2009;12 Suppl 1:12–17. [DOI] [PubMed] [Google Scholar]
  • 5. Shapiro S. Oral contraceptives, hormone therapy and cardiovascular risk. Climacteric. 2008;11(5):355–363. [DOI] [PubMed] [Google Scholar]
  • 6. Jackson SP. The growing complexity of platelet aggregation. Blood. 2007;109(12):5087–5095. [DOI] [PubMed] [Google Scholar]
  • 7. Ho-Tin-Noe B, Demers M, Wagner DD. How platelets safeguard vascular integrity. J Thromb Haemost. 2011;9 Suppl 1:56–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Ascenzi P, Bocedi A, Marino M. Structure-function relationship of estrogen receptor alpha and beta: Impact on human health. Mol Aspects Med. 2006;27(4):299–402. [DOI] [PubMed] [Google Scholar]
  • 9. Watson CS, Bulayeva NN, Wozniak AL, Finnerty CC. Signaling from the membrane via membrane estrogen receptor-alpha: Estrogens, xenoestrogens, and phytoestrogens. Steroids. 2005;70(5–7):364–371. [DOI] [PubMed] [Google Scholar]
  • 10. Ueda K, Karas RH. Emerging evidence of the importance of rapid, non-nuclear estrogen receptor signaling in the cardiovascular system. Steroids. 2013;78(6):589–596. [DOI] [PubMed] [Google Scholar]
  • 11. Vasudevan N, Kow LM, Pfaff DW. Early membrane estrogenic effects required for full expression of slower genomic actions in a nerve cell line. Proc Natl Acad Sci U S A. 2001;98(21):12267–12271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Temple JL, Wray S. Bovine serum albumin-estrogen compounds differentially alter gonadotropin-releasing hormone-1 neuronal activity. Endocrinology. 2005;146(2):558–563. [DOI] [PubMed] [Google Scholar]
  • 13. Han HJ, Lee YH, Park SH. Estradiol-17beta-BSA stimulates Ca(2+) uptake through nongenomic pathways in primary rabbit kidney proximal tubule cells: Involvement of cAMP and PKC. J Cell Physiol. 2000;183(1):37–44. [DOI] [PubMed] [Google Scholar]
  • 14. Stevis PE, Deecher DC, Suhadolnik L, Mallis LM, Frail DE. Differential effects of estradiol and estradiol-BSA conjugates. Endocrinology. 1999;140(11):5455–5458. [DOI] [PubMed] [Google Scholar]
  • 15. Harrington WR, Kim SH, Funk CC, et al. Estrogen dendrimer conjugates that preferentially activate extranuclear, nongenomic versus genomic pathways of estrogen action. Mol Endocrinol. 2006;20(3):491–502. [DOI] [PubMed] [Google Scholar]
  • 16. Kim SH, Katzenellenbogen JA. Hormone-PAMAM dendrimer conjugates: Polymer dynamics and tether structure affect ligand access to receptors. Angew Chem Int Ed Engl. 2006;45(43):7243–7248. [DOI] [PubMed] [Google Scholar]
  • 17. Anstead GM, Carlson KE, Katzenellenbogen JA. The estradiol pharmacophore: Ligand structure-estrogen receptor binding affinity relationships and a model for the receptor binding site. Steroids. 1997;62(3):268–303. [DOI] [PubMed] [Google Scholar]
  • 18. Chambliss KL, Wu Q, Oltmann S, et al. Non-nuclear estrogen receptor alpha signaling promotes cardiovascular protection but not uterine or breast cancer growth in mice. J Clin Invest. 2010;120(7):2319–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Adlanmerini M, Solinhac R, Abot A, et al. Mutation of the palmitoylation site of estrogen receptor α in vivo reveals tissue-specific roles for membrane versus nuclear actions. Proc Natl Acad Sci U S A. 2013;111(2):E283–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Arao Y, Hamilton KJ, Ray MK, Scott G, Mishina Y, Korach KS. Estrogen receptor α AF-2 mutation results in antagonist reversal and reveals tissue selective function of estrogen receptor modulators. Proc Natl Acad Sci U S A. 2011;108(36):14986–14991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Billon-Gales A, Krust A, Fontaine C, et al. Activation function 2 (AF2) of estrogen receptor-alpha is required for the atheroprotective action of estradiol but not to accelerate endothelial healing. Proc Natl Acad Sci U S A. 2011;108(32):13311–13316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Valera MC, Gratacap MP, Gourdy P, et al. Chronic estradiol treatment reduces platelet responses and protects mice from thromboembolism through the hematopoietic estrogen receptor α. Blood. 2012;120(8):1703–1712. [DOI] [PubMed] [Google Scholar]
  • 23. Elhage R, Arnal JF, Pieraggi MT, et al. 17 beta-estradiol prevents fatty streak formation in apolipoprotein E-deficient mice. Arterioscler Thromb Vasc Biol. 1997;17(11):2679–2684. [DOI] [PubMed] [Google Scholar]
  • 24. Jänne M, Hogeveen KN, Deol HK, Hammond GL. Expression and regulation of human sex hormone-binding globulin transgenes in mice during development. Endocrinology. 1999;140(9):4166–4174. [DOI] [PubMed] [Google Scholar]
  • 25. Séverin S, Gratacap MP, Lenain N, et al. Deficiency of Src homology 2 domain-containing inositol 5-phosphatase 1 affects platelet responses and thrombus growth. J Clin Invest. 2007;117(4):944–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Laurent PA, Séverin S, Hechler B, Vanhaesebroeck B, Payrastre B, Gratacap MP. Platelet PI3Kβ and GSK3 regulate thrombus stability at a high shear rate. Blood. 2015;125(5):881–888. [DOI] [PubMed] [Google Scholar]
  • 27. de Witt SM, Swieringa F, Cavill R, et al. Identification of platelet function defects by multi-parameter assessment of thrombus formation. Nat Commun. 2014;5:4257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Abot A, Fontaine C, Buscato M, et al. The uterine and vascular actions of estetrol delineate a distinctive profile of estrogen receptor α modulation, uncoupling nuclear and membrane activation. EMBO Mol Med. 2014;6(10):1328–1346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Abot A, Fontaine C, Raymond-Letron I, et al. The AF-1 activation function of estrogen receptor α is necessary and sufficient for uterine epithelial cell proliferation in vivo. Endocrinology. 2013;154(6):2222–2233. [DOI] [PubMed] [Google Scholar]
  • 30. Bernelot Moens SJ, Schnitzler GR, Nickerson M, et al. Rapid estrogen receptor signaling is essential for the protective effects of estrogen against vascular injury. Circulation. 2012;126(16):1993–2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Wyckoff MH, Chambliss KL, Mineo C, et al. Plasma membrane estrogen receptors are coupled to endothelial nitric-oxide synthase through Galpha(i). J Biol Chem. 2001;276(29):27071–27076. [DOI] [PubMed] [Google Scholar]
  • 32. Le Romancer M, Treilleux I, Leconte N, et al. Regulation of estrogen rapid signaling through arginine methylation by PRMT1. Mol Cell. 2008;31(2):212–221. [DOI] [PubMed] [Google Scholar]

Articles from Endocrinology are provided here courtesy of The Endocrine Society

RESOURCES