INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

PROLONGED ESTROGEN replacement therapy, such as that prescribed to postmenopausal women, not only is protective against cardiovascular disease (27) but also results in decreases in blood pressure (1) and increases in blood volume (BV), heart rate (HR), and cardiac output (CO) (18, 22, 39, 40). In addition, the cardiovascular alterations observed during normal pregnancy include substantial elevations in uteroplacental blood flow (20, 30-32, 34, 35), CO (20, 30, 31), and BV (18, 20), as well as decreases in mean arterial pressure (MAP) and systemic vascular resistance (SVR) (20, 30). Because plasma estrogen concentrations increase dramatically during gestation (6, 20, 25, 35), the prolonged exposure of the cardiovascular system to estrogen has been postulated to in part mediate and/or maintain these hemodynamic changes (18, 20, 22-24). This hypothesis is supported by observations in the ovine model that acute responses following a systemic bolus dose of 17-estradiol (E₂) (1 µg/kg) include decreases in uterine vascular resistance (UVR) and SVR and increases in CO by 90-120 min, whereas MAP essentially remains unchanged (9, 15, 22-24, 31, 33, 34). When regional tissue blood flows were examined in these studies, acute E₂-induced vasodilation (120 min) occurred in both reproductive and systemic (nonreproductive) tissues, e.g., skin and heart (31, 33, 34).

Recently we reported in nonpregnant sheep the effects of prolonged administration of "physiological" doses of E₂ that were chosen to achieve ovine gestation levels of estrogen. Although this dose may be somewhat lower than those used in postmenopausal E₂ replacement therapy, it caused a progressive rise in CO and this increase in systemic flow occurred independently of changes in uterine blood flow (UBF) (22). It is unknown, however, which systemic vascular beds increase flow during the prolonged administration of E₂. The majority of studies that have evaluated estrogen-mediated changes in regional blood flows using radiolabeled microspheres have been short-term studies, i.e., ~2-h responses after exogenous E₂ administration. However, because estrogen exposure with postmenopausal estrogen replacement therapy and during gestation is prolonged, an important issue that has not been addressed previously is whether differences in regional blood flow responses to acute versus prolonged exogenous estrogen treatment relate to the "duration of estrogen exposure." These data not only will provide the first direct physiological information as to which reproductive and nonreproductive vascular beds are responsive to prolonged estrogen therapy but also will address which regional changes in blood flows are responsible for the prolonged E₂-induced increases in CO and decreases in SVR (22).

In the present study, we tested the hypothesis that prolonged administration of a physiological dose of E₂ to ovariectomized nonpregnant ewes (22) will increase systemic perfusion by redistributing this flow to the skin, heart, skeletal muscle, and possibly several other nonreproductive tissues that are associated with decreases in SVR. To better understand the relationships between uterine and systemic blood flow responses to prolonged estrogen exposure and to contrast this to our previous studies (22, 24) of acute and prolonged estrogen administration, we determined 1) whether the overall reproductive and nonreproductive tissue vascular resistances and blood flows respond similarly to acute versus prolonged E₂ treatment; 2) whether prolonged E₂ treatment causes sustained decreases in vascular resistance and increases in skin, renal, coronary, brain, and skeletal muscle blood flows, independent of changes in UBF; 3) whether the changes in UBF responses are similar in the endometrium, myometrium, and caruncles versus the other nonuterine reproductive vascular beds; 4) whether the changes in skin, renal, coronary, and brain blood flows show regional differences in sensitivity; and 5) which regional systemic vascular beds exhibit increases in blood flow to account for the E₂-induced rise in CO and decreases in SVR.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Surgical preparation. The following protocols were similar to our previous studies (22, 24) and were approved by the Research Animal Care and Use Committee of the University of Wisconsin Colleges of Medicine and Agriculture and Life Science. Fifteen nonpregnant mixed-breed (68 ± 3.6 kg) ewes were used in this study. Twenty-four hours before surgery, the animals were fasted and water was removed. On the day of surgery, animals were given an intramuscular injection of ketamine (15 mg/kg), atropine (12 µg/kg), and antibiotics (400,000 U penicillin, 0.5 g streptomycin). A catheter was inserted into the jugular vein via a percutaneous needle (14 gauge) and was advanced into the right ventricle. An adequate plane of surgical anesthesia was obtained with a continuous intravenous drip of 10% ketamine (10 mg/ml) in 0.9% saline and 5% dextrose with pentobarbital sodium (50 mg/ml) supplemented as needed for additional analgesia based on the state of arousal of the ewe. The abdominal, inguinal, and cervical regions were then shaved and aseptically scrubbed. Under standard sterile conditions, a midventral laparotomy was performed, the uterus was exposed, and the ovaries were extirpated to remove the source of cyclic secretion of estrogen and progesterone. Electromagnetic (model RC-2000 monitor, Micron Instruments, Los Angeles, CA) flow probes (n = 2 E₂-treated animals) or Transonic (model T101 monitor, Ithaca, NY) flow probes (n = 13 animals) were placed around both uterine arteries proximal to the first bifurcation in the mesometrium. After closure of the midline incision, arterial and venous catheters (polyvinyl; 0.032-in ID, 0.051-in OD) were inserted into both superficial saphenous femoral vessels and advanced through the femoral circulation into the descending aorta (10 cm) and abdominal inferior vena cava (10 cm). All catheters were filled with heparinized saline (100 IU/ml), sealed, and exteriorized with the flow probe leads to a pouch on the ewe's flank. A catheter was then inserted into the left ventricle via the right carotid artery, and its position was confirmed by the characteristic pressure pattern recorded on a computer through a DATAQ (Akron, OH) data acquisition system. The ewe was then returned to the holding pens and given free access to food and water ad libitum. Intramuscular antibiotic treatment (250,000 U penicillin, 0.3 g streptomycin) was given for the next 5 days postsurgery and every third day thereafter.

Experimental protocol. Animals were returned to the laboratory on the fifth day postsurgery. On the following day, after a 60-min control period, a 1 µg/kg intravenous injection of E₂ was given and UBF, HR, and MAP were monitored for 2 h to ensure a normal acute uterine vasodilatory response as described previously (15, 22-24, 33). This procedure was repeated on the next day and compared with the first response to ensure that all animals studied responded similarly on both days, thus indicating the recovery from postsurgical stresses.

1

1

1

1

Hemodynamic measurements. Cardiovascular parameters (MAP, HR, and UBF) were recorded during each daily monitoring period (45-60 min) on an IBM PC computer through a DATAQ data acquisition system and stored. Six to ten 1-min segments over ~10-min intervals every 15 min were averaged for MAP, HR, and UBF on days when the ewes were monitored without injection of microspheres and on days 3, 6, 8, and 10 before and during each microsphere injection. Regional blood flows were calculated using the standard radioactive microsphere technique for each organ by the following equation: withdrawal rate divided by number of microspheres in reference blood times number of microspheres in tissue samples. Only those tissues that had >400 microspheres, determined for all six microsphere labels for each tissue evaluated, were considered valid and are reported. Resistance was calculated as MAP/regional blood flow. In preliminary studies, we determined that the skin blood flow response to E₂ varied according to the site of sampling; therefore, skin blood flows were derived from 10 different sites (see Experimental protocol) to allow for regional changes from areas that may be more or less sensitive to E₂. Skeletal muscle blood flows were obtained from multiple sites, which did not differ in flow and were calculated as a percent muscle mass by weight for each ewe taken from published agricultural tables (37). CO was calculated as the number of microspheres injected divided by the number of microspheres in arterial reference sample times the withdrawal rate. SVR was obtained by dividing MAP by CO.

Statistical analysis. Data were analyzed by ANOVA or Student's t-tests where appropriate. Changes were considered statistically different at P < 0.05. All values are reported as means ± SE. Data were analyzed as absolute flows (ml/min) and percent change from control (%) for relative responses compared with their day 0 controls or between acute (120 min) and prolonged (days 3-10) treatments.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Cardiovascular effects of acute vs. prolonged E₂ infusion. Animals were given 1 µg/kg E₂ on the sixth and seventh days postsurgery and had the expected standard and reproducible estrogen vasodilatory responses to E₂ (15, 22-24), i.e., increases in UBF ( 212 ± 19 ml/min; 900-1,200%), no change in MAP, and 20-30% increases in HR. An estrogen withdrawal period of 4 days then ensued, during which time no appreciable hemodynamic changes occurred (data not shown). Control values for hemodynamic parameters (day 0) were similar to those reported previously for nonpregnant ovariectomized sheep (9, 22, 24, 41): MAP = 87 ± 5 mmHg; HR = 83 ± 5 beats/min; UBF = 23 ± 7 ml/min. The acute (120 min) effects of the 5 µg/kg loading dose of E₂ given on day 0 were similar to the responses to a 1 µg/kg dose given on postoperative days 6 and 7, i.e., MAP was not changed, HR increased (P < 0.01) by 31 ± 6%, and UBF determined by the flow probe technique increased from 23 ± 7 to 318 ± 48 ml/min with a concomitant decrease in UVR (Fig. 1). E₂-mediated decreases (P < 0.01) in SVR from 12.4 ± 1.3 to 10.1 ± 0.9 mmHg · min · l¹ in association with increases in CO from 7.51 ± 0.94 to 9.12 ± 0.76 l/min also were observed. Acute treatment with vehicle did not alter these cardiovascular parameters (Fig. 1). We then contrasted these effects of acute vehicle and E₂ injection with the prolonged treatments (Fig. 1 vs. Fig. 2). Although MAP was unaltered by acute vehicle or estrogen treatment and prolonged vehicle administration, it was significantly decreased after 3-4 days of prolonged E₂ treatment, with hypotension of 3-9% being observed through day 8. In contrast to vehicle infusions, HR values were increased (P < 0.05) by both acute and prolonged estrogen treatments, with increases ranging from 4 to 31%; however, this was variable after day 4. Measurements of UBF using the flow probes exhibited initial dramatic increases (P < 0.001) within 2 h of E₂ but not vehicle infusion, averaging 1,810 ± 270%, and UVR decreased 93 ± 1.3% (Fig. 1); however, E₂-induced uterine responses were not sustained at these elevated levels throughout the duration of estrogen treatment (Fig. 2), confirming previous observations from this (22) and other (7) laboratories. As an index of the stability of the animal preparation during infusion of E₂, arterial blood gases (pH, PCO₂, PO₂, [O₂], O₂ saturation), hemoglobin, glucose, lactate, and hematocrit were obtained. There were no significant (P < 0.05) differences between day 0 controls versus the values observed throughout the E₂ infusion for pH (7.47 ± 0.03 vs. 7.43 ± 0.01), PCO₂ (43.6 ± 0.9 vs. 40.8 ± 0.5 mmHg), PO₂ (97 ± 2 vs. 101 ± 5 mmHg), [O₂] (11.94 ± 1.47 vs. 11.7 ± 1.7 vol%), O₂ saturation (95.8 ± 2.46 vs. 94.6 ± 4.9%), hemoglobin (9.33 ± 0.87 vs. 9.36 ± 0.90%), glucose (38.3 ± 9.0 vs. 33.6 ± 8.1 mg/dl), lactate (0.97 ± 0.30 vs. 1.05 ± 0.38 mmol/l), or hematocrit (28.2 ± 1.9 vs. 27.0 ± 2.0%).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Acute (120 min) effects of vehicle treatment (n = 6) or 17-estradiol (E2; 5 µg/kg) treatment (n = 9) on systemic [mean arterial pressure (MAP), heart rate (HR); A] and uterine [uterine blood flow (UBF), uterine vascular resistance (UVR); B] vascular responses in nonpregnant ovariectomized sheep. Control, combined pretreatment controls (n = 15) before injection of vehicle or E2 treatment. Values are means ± SE; ** P < 0.01 vs. pretreatment control and vehicle.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Effects of prolonged vehicle treatment (n = 6) or E2 (6 µg/kg over 24 h for 10 days) treatment (n = 9) on daily systemic [MAP (A) and HR (B)] and uterine [UBF (C) and UVR (D)] vascular responses in nonpregnant ovariectomized sheep. Daily UBF was measured using flow probe technique. Prolonged E2 infusion was immediately preceded by a 5 µg/kg bolus loading dose of E2. Values are means ± SE. Significant differences vs. day 0 pretreatment control: + 0.05  P < 0.1, * P < 0.05, ** P < 0.01, *** P < 0.001.

Effects of E₂ on blood flows in reproductive vs. nonreproductive tissues. Total reproductive blood flow (ml/min) was calculated as the summation of flows to the uterus (endometrium, myometrium, and caruncles), oviducts, broad ligament, cervix, vagina, vulva, and mammary gland as defined previously (31, 33, 34), whereas nonreproductive (systemic) blood flow was the summation of flows to the remaining tissues evaluated in this study (see MATERIALS AND METHODS). Total reproductive blood flow exhibited significant increases in response to E₂ throughout the study period when compared with control; however, because responses were substantially lower after the 120-min E₂-induced peak, tachyphylaxis to estrogen appears to have occurred, although estrogen receptor internalization is another likely explanation (Fig. 3). Nonreproductive blood flow responded to E₂ with an overall increase in systemic flows that did not reach statistical significance (P < 0.08) during the acute estrogen phase but did achieve significance on days 3, 6, 8, and 10 during the prolonged infusion of E₂. It is noteworthy that the y-axes on Fig. 3, A and B, are one order of magnitude different, such that even very small changes in nonreproductive blood flow greatly exceed those noted in reproductive tissues on an absolute milliliter-per-minute basis. By contrast, when the relative response (% from baseline) in reproductive and nonreproductive blood flows was evaluated (Fig. 3C), as an indicator of vascular sensitivity to the vasodilatory effects of E₂, the percent change in reproductive blood flows was considerably greater (P < 0.0001) than that observed for the nonreproductive tissues. Moreover, on a relative basis, the percent change in nonreproductive blood flows was statistically significant (P < 0.05) at 120 min after the acute estrogen treatment and on days 3, 6, 8, and 10 of the E₂ infusion. Infusion of vehicle in six sheep did not appreciably alter (P > 0.05) reproductive or nonreproductive tissue blood flows, blood gases, lactate, glucose, or hematocrit (data not shown).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Total reproductive (A) and nonreproductive (B) tissue blood flows (ml/min) and relative changes (C) in blood flows (% from control) during acute (5 µg/kg; 120 min) and prolonged (6 µg/kg over 24 h for 10 days) E2 administration to nonpregnant ovariectomized sheep (n = 9). Prolonged E2 infusion was immediately preceded by 5 µg/kg bolus loading dose of E2. Pretreatment control values are indicated by open bars and symbols; acute and prolonged effects of E2 are shown with filled and shaded bars and symbols, respectively. Absolute changes in reproductive blood flow during prolonged E2 treatment were less than nonreproductive blood flows (P < 0.001). Vehicle-treated animals did not show significant changes in either reproductive or nonreproductive blood flows (data not shown). Values are means ± SE. + 0.05  P < 0.1; * P < 0.05; ** P < 0.01; *** P < 0.001, significant differences versus day 0 pretreatment control.  P < 0.01, significant differences of acute vs. prolonged E2.

Changes in regional reproductive and nonreproductive vascular bed resistances and blood flows. Vascular resistances and absolute blood flows of the individual reproductive tissues were inversely related (Table 1). Comparisons of UVR and UBF measured by the flow probe (Figs. 1 and 2) and microsphere (Table 1) techniques showed parallel changes. Vehicle infusion did not significantly alter either reproductive tissue vascular resistance or blood flows (Table 1). Although decreases in vascular resistance and increases in flow to all three of the tissues comprising the uterus measured during the acute estrogen response were very dramatic, there was a considerable loss of the vasodilatory response with prolonged E₂ infusion. This initial dramatic vasodilatory response followed by "tachyphylaxis" was similar in most of the individual reproductive tissues studied except the vulva and mammary, which showed elevated blood flows throughout the prolonged E₂ infusion. When comparing relative blood flow responses (%) of the endometrium, myometrium, and caruncles (Fig. 4), we observed that there were parallel E₂-induced changes in perfusion but no effect of vehicle infusion. The intrauterine sensitivity of the endometrium, myometrium, and caruncles also was evaluated as the percentage of total UBF (Ref. 23; Fig. 4). It was observed that each of these tissues comprised ~25-40% of UBF and only exhibited minor changes with acute E₂ treatment. During prolonged E₂ treatment, however, blood flow to myometrium increased to ~50% of UBF at the expense of caruncular flow. Vehicle infusion did not alter the distribution of blood flow within the uterus. We also observed that the percent change from control blood flow to the cervix, oviduct, vagina, and broad ligament exhibited a similar pattern of flow as noted for the uterus, whereas the relative percent response of the vulva and mammary gland did not show significant decreases in prolonged compared with 120 min of acute estrogen (Fig. 5). The acute relative E₂-induced increase in vulva, broad ligament, and mammary gland blood flows were nearly one-half the percent change of the uterine, cervical, oviductal, and vaginal vascular beds.

View this table: [in this window] [in a new window]   Table 1.   Reproductive tissue vascular resistance and blood flow during vehicle infusion and during acute and prolonged E2 administration to nonpregnant ovariectomized sheep

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Relative changes (A and B; % from control) and distribution (C and D; % of total UBF) of UBF during vehicle (n = 6) infusion (A and C) and during acute (5 µg/kg; 120 min) and prolonged (6 µg/kg over 24 h for 10 days) E2 (n = 9) administration (B and D) to nonpregnant ovariectomized sheep. Prolonged E2 infusion was immediately preceded by 5 µg/kg bolus loading dose of E2. Total UBF is sum of blood flows measured by microsphere technique to caruncles, endometrium, and myometrium. Vehicle infusion did not significantly alter either UBF or its distribution (n = 6). Values are means ± SE. Significant differences vs. day 0 pretreatment control: + 0.05  P < 0.1; * P < 0.05; ** P < 0.01; *** P < 0.001.  P < 0.01, significant differences of acute vs. prolonged E2.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Relative changes (% from control) in nonuterine individual reproductive vascular bed blood flows during vehicle infusion (A and C) and during acute (5 µg/kg; 120 min) and prolonged (6 µg/kg over 24 h for 10 days) E2 (n = 9) administration (B and D) to nonpregnant ovariectomized sheep. Prolonged E2 infusion was immediately preceded by 5 µg/kg bolus loading dose of E2. The 3 tissues shown in A and B responded similarly. Vulva and mammary gland blood flows were elevated throughout E2 infusion period. Vehicle infusion did not significantly alter blood flows to nonuterine reproductive vascular bed. Values are means ± SE. Significant differences vs. day 0 pretreatment control: * P < 0.05, ** P < 0.01, *** P < 0.001.  P < 0.001, significant differences of acute vs. prolonged E2.

View this table: [in this window] [in a new window]   Table 2.   Nonreproductive tissue resistance during vehicle infusion and during acute and prolonged E2 administration to nonpregnant ovariectomized sheep

View this table: [in this window] [in a new window]   Table 3.   Nonreproductive tissue blood flow during vehicle infusion and during acute and prolonged E2 administration to nonpregnant ovariectomized sheep

View this table: [in this window] [in a new window]   Table 4.   Relative changes in nonreproductive tissue blood flow to organs that contribute substantially to cardiac output during vehicle infusion and during acute and prolonged E2 administration to nonpregnant ovariectomized sheep

Regional sensitivity of skin, brain, coronary, and kidney blood flow distribution. Increases in blood flow within different areas of an organ are a reflection of the regional sensitivity to vasoactive agents such as E₂. Blood flow distribution to regions of the skin during E₂, but not vehicle, treatment acutely demonstrated increases in nearly all regions (Table 5). The relative percent change from baseline blood flow responses of the skin of the face, mammary region, and vulva showed very high sensitivity and continued progressive increases with prolonged E₂ treatment (data not shown). Blood flow to the four regions of the brain demonstrated similar increases in perfusion, but these increases in flow were only observed during the prolonged phase of the E₂ infusion. Moreover, both the atria versus ventricle as well as the renal cortex versus medulla appeared to respond similarly to E₂ infusion (Table 5). Infusion of vehicle was without effect on the regional change in skin, brain, coronary, or renal blood flows, although there was a trend for flows to increase in cerebrum, midbrain, and ventricles.

View this table: [in this window] [in a new window]   Table 5.   Skin, brain, coronary, and kidney blood flow distribution during vehicle infusion and during acute and prolonged E2 administration to nonpregnant ovariectomized sheep

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Hemodynamic responses observed during prolonged estrogen administration are of substantial clinical importance because hormone replacement therapy in postmenopausal women is beneficial for the prevention of cardiovascular disease (1, 27). This model also mimics many of the systemic cardiovascular adaptations of pregnancy (22), suggesting that placental production of estrogens (6, 25, 35) may play a role in the cardiovascular adaptation and fluid volume changes that normally occur (18, 20, 39, 40). However, because both estrogen and progesterone are produced by the placenta (25, 35), they both may be needed to fully reproduce these hemodynamic changes. Previously, we reported that prolonged E₂ treatment persistently decreases MAP and SVR while increasing CO, HR, stroke volume, and BV; however, substantial changes in UBF and UVR were not maintained. We have extended these observations by specifically showing which vascular beds have decreases in vascular resistance, thus defining which tissue blood flows account for the increases in CO during prolonged estrogen treatment. We present here the first detailed description of the time course of the E₂-mediated changes in blood flows to specific reproductive and nonreproductive vascular beds as well as the distribution of flow within the uterus, skin, brain, heart, and kidney. In the current and most previous studies of this nature (15, 22, 24, 33, 39-41) the ewes were ovariectomized to remove the endogenous source of cyclical estrogen and progesterone. Therefore, it is possible that the decrease in ovarian steroids will potentiate the observed vascular responses to E₂. Although this remains to be proven, this premise is not supported by the observations that the vehicle-treated animals did not have a continued fall in blood flow to either the reproductive or nonreproductive tissues throughout the 10-day infusion period. These blood flow data as well as the blood gas, glucose, and lactate analysis also provide evidence that multiple microsphere injections can indeed be administered without compromising the stability of this chronic physiological animal preparation.

A single intravenous bolus of E₂ results in progressive 1- to 2-h, but transient (<24 h), decreases in UVR and SVR with increases in UBF and CO, although acute administration of estrogen does not decrease MAP (Fig. 1; Refs. 9, 22-24, 32). In the present study using 5 µg E₂/kg, we confirmed the finding in previous studies using 1 µg E₂/kg that this acute (~2 h) vasodilatory response to E₂ is associated with similar maximal increases in blood flow to the reproductive tissues (e.g., cervix, uterus, vagina, etc.), bladder, heart, and skin (31, 33, 34). We also report here minor acute E₂-induced rises in blood flow to the esophagus, adrenals, kidney, and trachea (Table 3). Prolonged E₂ treatment results in persistent and progressive systemic vasodilation, with increases in CO of as much as 2-3 l/min (22); however, the destination of this rise in CO has never been described. It is obvious from our previous and current studies that the uterus contributes only somewhat to the rise in CO and that mainly in the first 2-3 days of estrogen treatment were there substantial changes in UBF. Previously, we hypothesized (22, 24) that the skin blood flow, which represents the largest organ in the body, may be responsible for the majority of the elevation in systemic blood flow. This was only in part proven in the current study and is consistent with the observations of increased CO and blood flow through the extremities during estrogen replacement therapy (5, 14, 22). Counter to our previous suggestion, we observed that coronary blood flow can indeed account for a portion (5-14%) of the rise in CO but noted that with prolonged but not acute E₂ treatment the skeletal muscle, brain, spleen, and pancreas also collectively play a major role in accounting for increases in systemic blood flow (CO). It is noteworthy that the changes in skeletal muscle blood flow directly relate to the anabolic effects of estrogen reported in the agricultural industry with increases in weight gain during the use of estrogenic compounds such as diethylstilbestrol for growth promotion in beef cattle (12).

Although with prolonged E₂ treatment MAP (perfusion pressure) decreased modestly (3-9%), no animals became severely hypotensive. Partial restoration of the "underfilled state" caused by the systemic vasodilation occurs via increases in BV, which directly impact on the maintenance of MAP (18, 20, 22, 39). This rise in BV therefore also partially accounts for the gradual but continued maintenance and/or increases in blood flow in certain nonreproductive vascular beds (e.g., skin, skeletal muscle, etc.). It is expected that as CO was redistributed to these major tissues that control SVR, blood flow would be observed to decrease in other vascular beds if "underfilling" persists. This was not the case, because the majority of flows to nonreproductive vascular beds actually increased and most others stayed approximately the same, suggesting that the rise in BV (18, 20, 22, 39) was important in the restoration and/or redistribution of blood flows to these critical vascular beds. Our studies were not intended to address the direct effects of E₂ on adrenal aldosterone secretion or regional changes in flow of the adrenal cortex; however, the whole adrenal did show marginal acute increases in flow in a fashion similar to the kidney. In previous studies others did not observe increases in adrenal (33) or renal (28, 33) blood flows with E₂ treatments, which may be related to the higher dose of E₂ we administered.

On a relative basis (%), reproductive tissues had a much greater response to E₂ than nonreproductive tissues (Fig. 3C), suggesting that it has importance in terms of perfusion in those tissues exposed to high local concentrations of steroids derived from the ovary and/or placenta (21, 25, 35). However, with regard to elevations in CO and BV (22) and the much greater absolute elevations in blood flows to nonreproductive versus reproductive vascular beds, the latter is of less consequence. This derives from the current observations that increases in total blood flows to the reproductive vascular bed (uterine, oviductal, broad ligament, cervical, etc.) account for the minority (3-17%) of the changes in CO (22) with prolonged estrogen treatment. The bladder appears to have been incorrectly classified with the nonreproductive vascular bed (31, 33, 34), because it responded similarly to the reproductive vascular beds. The urinary tract has estrogen receptors, and bladder function is, in part, maintained by steroid hormones such as estrogen and progesterone (3, 36). Increase in urinary stress incontinence is observed in postmenopausal women, which is alleviated by treatment with estrogen replacement therapy.

Increases in skin blood flow confirm studies in postmenopausal women on estrogen replacement therapy (5), but these were regionalized such that blood flow to the pigmented area covering the vulva, i.e., the "sex skin," as well as to the skin overlying the mammary gland and face was much more responsive on a % basis to estrogen treatment than blood flows to the extremities, back, or abdomen. This observation relates teleologically to that noted in certain nonhuman primates where the sex skin of the baboon or monkey becomes red, edematous, and swollen near the time of ovulation in these species when estrogen is elevated (45). This change in blood flow in these primates is accompanied by changes in coloration due to hyperemia, thus providing visual cues that have evolved in species that do not strictly exhibit behavioral estrus. Moreover, postmenopausal women exhibit atrophic external genitalia, a decrease in the integrity and tone of their entire skin, as well as other mucous membrane-related tissues, e.g., the atrophic vaginal epithelium. It is of further interest that E₂-induced increases in skin blood flow were observed to occur in numerous regions. It is likely that with increases in CO (22), oxygen consumption (10), and thus basal metabolic rate, elevations in skin blood flow that occur with estrogen replacement therapy and with pregnancy are mechanisms that have evolved to dissipate heat.

We have confirmed that prolonged E₂ treatment increases HR and CO and decreases MAP (11, 16, 18, 22, 39, 40); however, HR and CO responses in our current study were somewhat more variable, increasing from 4 to 31% with inconsistent changes after day 4. One possible explanation for our variable results in HR and CO relates to the fact that, unlike previous studies, we chronically implanted catheters in both the right and left ventricles for measurements of mixed venous blood gases and infusion of microspheres, respectively. Moreover, when measuring CO by the microsphere method, we can only obtain one observation per day whereas the dye dilution technique we used previously (22, 24) entailed averaging three to four measurements per day and therefore had much greater precision. Because E₂ has been shown to accumulate in the nuclei of atrial myocytes, to cause negative inotropism (29) as well as positive chronotropism in isolated perfused rabbit hearts (2), the rise in CO with E₂ treatment may relate to direct effects on the heart.

We have demonstrated that acute and prolonged administration of estrogen increases coronary blood flow as reported previously for acute treatments (16, 17, 31, 33, 34). We report for the first time here that these increases in coronary blood flow occur proportionally in both the atria and ventricles (Table 5). We initially observed a 77% increase in total coronary blood flow that remained elevated above control throughout the E₂ infusion at levels between 32 and 190%. Increases in cardiac perfusion on a relative % basis appear to decline somewhat from day 3 to day 10; however, because CO also increases with prolonged E₂ administration (11, 22, 40), the efficiency of the cardiac function may be improved. Therefore, the increase in left ventricular chamber enlargement and end-diastolic volumes with estrogen replacement therapy (11) may require the rise in total coronary blood flow for normal and efficient function. To our knowledge, it is unknown whether estrogen increases the efficiency of cardiac function while it increases cardiac workload. The cellular and molecular mechanisms by which estrogen increases cardiac blood flow appear to be related to an increase in nitric oxide (14, 16) as well as endothelial nitric oxide synthase (eNOS) expression in coronary artery endothelium (17) or to a direct effect of estrogen on the coronary artery vascular smooth muscle NOS (8, 44).

Brain blood flow progressively increased from day 3 to day 10 of estrogen exposure. Initially, decreases in MAP (2 days) and thus perfusion pressure may have affected brain blood flow through autoregulatory mechanisms. However, because MAP did not decrease further with prolonged E₂ administration, it is likely that, in contrast to other nonreproductive vascular beds in which blood flows increased acutely and were maintained (e.g., coronary, skin, and esophagus), this late phenomenon in the brain may be a unique observation related to cerebral vasodilation. The additional increases in brain blood flow we observed on days 8 and 10 of the E₂ infusion may relate to increased mental performance, brain activity, cognitive functioning, and improvement in cerebral blood flow, the latter measured by positron emission computer tomography scans, during estrogen replacement therapy in women (4, 13, 38). Moreover, by evaluating the relative changes in the regional distribution within the brain (Table 5), it was noted that all of the regions of the brain responded in a similar fashion.

The mechanism by which estrogen causes either acute or prolonged systemic vasodilation is still not understood and may reflect increases in endothelium-derived vasodilators [e.g., nitric oxide (14) or prostacyclin] or in the alterations and remodeling of vascular smooth muscle. We and others reported that pretreatment of the uterus with the NOS inhibitor N-nitro-L-arginine methyl ester (L-NAME) attenuated the acute E₂-mediated vasodilation response (32, 41), whereas indomethacin had no effect (32). Lang et al. (16) reported that L-NAME will completely inhibit the acute estrogen-induced increase in coronary blood flow. It is not known whether L-NAME treatment of chronically estrogenized sheep will reverse the cardiovascular and regional tissue blood flow changes we observed. Furthermore, acute estrogen treatment increases cGMP, the second messenger for nitric oxide, in the uterine venous and systemic blood (26, 32), accounting for an increase in uterine cGMP secretion. There are also specific increases in uterine artery eNOS expression (calcium-dependent constitutive isoform) in response to acute (26) and prolonged (unpublished observations) estrogen treatment. Expression data for eNOS mRNA with prolonged in vivo estrogen exposure studies are not yet available; however, Veille et al. (42) reported that 3-day E₂ infusion into sheep increases calcium-dependent NOS specific activity in uterine but not renal arteries. Weiner et al. (43) also reported in the guinea pig increases in constitutive NOS activity in skeletal muscle, heart, and brain during prolonged E₂ treatment, which is consistent with the increases in blood flow we observed in these tissues (Table 3). Recently, MacRitchie et al. (19) demonstrated that in ovine fetal pulmonary artery endothelial cells, prolonged (24-48 h) in vitro treatment with E₂ increases eNOS mRNA and protein expressions.

In conclusion, we provide the first comprehensive and integrated description of the time course for the regional changes in vascular resistance and blood flow to both reproductive and nonreproductive vascular beds. We observed that the sensitivity of reproductive tissues to E₂ treatment on a relative basis is much greater than that of nonreproductive tissues. However, prolonged estrogen administration also maintains or further increases blood flows to several important nonreproductive vascular tissues, most notably the coronary, skin, skeletal muscle, brain, pancreas, and spleen vascular beds. Therefore, changes in blood flow to these vascular beds that make up a substantial portion of the mass of the body and thus CO are likely to be of great importance in modulating peripheral vascular resistance in women with postmenopausal estrogen replacement therapy and during pregnancy, although a modulation role for progesterone has yet to be elucidated. Because prolonged E₂ treatment is protective against cardiovascular disease in postmenopausal women (27), the further understanding of the integrative systemic and, more important, mechanistic regional hemodynamic and endocrine effects of this therapy is of substantial importance.