`Endocrinology
`Cr)pyrigh1,© 1997 by The Endocrine Society
`
`Vol, 138, No. 4
`Printed in U. S.A
`
`Characterization of the Ovariectomized Rat Model for
`
`the Evaluation of Estrogen Effects on Plasma Cholesterol
`Levels
`
`SCOTT G. LUNDEEN, JEFFREY M. CARVER, MAR—LEE MCKEAN, AND
`RICHARD C. WINNEKER
`
`Women’s Health Research Institute, Wyeth-Ayerst Research (S. G.L., J.M.C., R.C.W.), Radnor,
`Pennsylvania 19087; and Cardiovascular Research, Wyeth-Ayerst Research (M-L.M.}, Princeton,
`New Jersey 08852
`
`ABSTRACT
`Estrogens protect against cardiovascular disease in women
`through effects on the vascular wall and liver. Here we further char-
`acterize the rat as a model for the evaluation of estrogenic effects on
`plasma lipid levels us. uterine wet weight. In adult ovariectomized
`female rats treated for 4 days sc, 1701-ethi.nyl estradiol (EE) was the
`most potent agent to lower plasma total and high density lipoprotein
`cholesterol levels, followed by 1'73-estradiol and 1711-estradiol. How-
`ever, 17a-estradiol had the greatest separation of uterotropic vs. cho-
`lesterol-lowering effects. EE had the same lipid-lowering potency
`whether administered sc or orally to adult rats. It had no effect on
`cholesterol levels i.n immature rats, even though the uterotropic re-
`
`sponse was dramatic. Testosterone propionate, dexzunethasone, and
`progesterone did not significantly lower cholesterol levels. The a.n-
`tiestrogens tamoxifen and raloxifene lowered cholesterol levels, but
`with less efficacy and potency than the estrogens. ICI 182780 had no
`effect on cholesterol levels. When coadministered with EE, ICI 18 2780
`inhibited the cholesterol-lowering and uterotropic activities of EE,
`suggesting that the estrogen receptor pathway is involved. In con-
`clusion, although the information from the rat is limited as a model
`ofthe low density lipoprotein-lowering effects of estrogens in humans,
`it can be used to study the effects a.nd mechanism of action of estrogen
`and antiestmgens on plasma cholesterol levels. (Endocrinology 138:
`1552—1558, 1997)
`
`T HAS B4 4N recognized for many years that estrogens
`have a profound beneficial effect on the cardiovascular
`system in women (1). Women receiving hormone replace-
`ment therapy have approximately a 50% reduction in car-
`diovascular disease (2). Most of this evidence has come from
`clinical trials and studies with nonhuman primates that have
`clearly demonstrated the beneficial effect of estrogens (2-5).
`Such studies, however, have provided limited insight into
`the molecular mechanisms by which estrogens exert their
`beneficial effects. Recent studies have provided evidence for
`some of the potential targets through which estrogen may be
`protecting the cardiovascular system. It is now clear that this
`protection by estrogen is due to both direct effects on the
`blood vessel wall (6),
`i.e. through the regulation of anti-
`atherogenic agents such as nitric oxide and indirect effects in
`the liver. The end result of estrogen action in the liver is
`altered plasma cholesterol levels. In humans, estrogens de-
`crease circulating low density lipoproteins (LDL) and in-
`creases high density lipoprotein (HDL) (7-9).
`Rats have been used as a model system to study estrogenic
`effects on plasma lipid levels (10-12). The predominant
`plasma cholesterol in rats is HDL, not LDL, as it is i11 humans.
`Estrogens dramatically decrease both LDL and HDL choles-
`terol plasma levels in rats. Therefore, one acknowledged
`weakness of the rat model is that it is only useful for eval-
`
`Received September 20, 1996.
`Address all correspondence and requests for reprints to: Scott C.
`Lundeen, Ph.D., Women’; Healdi Research Institute, Wyeth-Ayerst Lab-
`oratories, l45 King of Prussia Road, Radnor, Pennsylvania 19087. E-mail
`address: lundees@war.wyethcom.
`
`uating the mechanisms involved in the LDL-lowering effects
`of estrogen and provides little, if any, relevant information
`on potential effects on HDL. As in humans, there is little
`informa ion on the molecular mechanism bywhich estrogens
`lower cholesterol in the rat. Early studies provided mecha-
`nistic clues as to how the estrogens mediate their effects on
`plasma ipids. It was shown that pharmacological doses of
`estrogens up-regulate LDL receptors in rat livers (13, 14) and
`in huma 1 hepatoma cell lines (15). It has also been shown that
`LDL binding in human liver homogenates is correlated with
`serum estrogen concentrations (16). Regulation of the LDL
`receptors has been shown to involve both transcriptional (17)
`and pos transcriptional (13, 14) mechanisms.
`There also are few data to support the role of the classical
`estrogen receptor (ER) pathway in mediating the lipid-low-
`ering effect of estrogens. Clearly, transcriptional regulation
`of the LDL receptor provides suggestive evidence for clas-
`sical ER control. However, there are few data to support this
`iypothesis, and direct evidence for ER involvement is still
`acking. In fact, there is evidence suggesting that a novel
`nechanism is involved. Firstly, the antiestrogens tamoxifen
`and raloxifene act as estrogen agonists in the liver, causing
`a decrease in total plasma cholesterol in rats and LDL in
`iumans (11, 18 -22). Secondly, the potencies of estrogens in
`he liver, as measured by changes in plasma cholesterol, do
`not correspond with their potencies in the uterus or their
`elative affinities for the ER (23).
`We initiated these studies to characterize the effects of
`
`estrogens on plasma lipid levels in rats as a model for the
`'ndirect cardioprotective effects of estrogen. In doing so, we
`iave examined several estrogenic and antiestrogenic com-
`
`1552
`
`Astrazeneca Ex. 2110 p. 1
`Mylan Pharms. Inc. v. Astrazeneca AB IPR2016-01324
`
`
`
`ESTROGEN EFFECTS ON PLASMA CHOLESTEROL LEVELS
`
`1553
`
`pounds in this system and studied the role of the ER in
`mediating the response in the liver vs. that in the uterus.
`Materials and Methods
`
`Reagents
`17a-Ethinyl estradiol (EE), 17&estradiol (17B—E2), 17a-estradiol (1711-
`E2), and dexamethasone were obtained From Sigma Chemical Co. (St.
`Louis, MO); tamoxifen citrate was obtained from Stuart Pharmaceuticals
`(Wilmington, DE); progesterone and testosterone propionate were ob-
`tained from Steraloids (I/Vilton, NH). ICI 182,780 was generously sup-
`plied by Zeneca Pharmaceuticals (Wilmington, DE). Raloxifene was
`synthesized by the Wyeth-Ayerst Medicinal Chemistry group. Stock
`solutions of the test compounds were prepared in either 100% ethanol
`or dimethylsulfoxide. The compounds were diluted into 10% ethanol in
`corn oil (Mazola, Best Food Division, CPC International Inc., Englewood
`Cliff, NI) vehicle before treatment of the anirnals.
`
`Animals and treatment protocols
`The research animals were housed in a facility accredited by the
`American Association for Accreditation of Laboratory Animal Care in
`accordance with the Animal Welfare Act and the Guide for the Care and
`Use of Laboratory Anirnals, and the study was approved by the insti-
`tutional animal care and use committee of VVyeth-Ayerst Research.
`Immature female (19 days old) or ovariectomized female (60 day-old)
`Sprague—Dawley rats were obtained from Taconic Farms (Germantown,
`NY). The ovariectomies were performed by the supplier a minimum of
`8 days before the first treatment. The animals were housed under a 12-h
`light, 12-h dark cycle and given Purina 5001 rodart chow (North Penn
`Feeds, North Wales, PA) and water ad l1lJii1iWI.. Upon arrival, the rats were
`randomized and placed in groups of four to eight, depending upon the
`experiment. The adult animals were given a minimum of 72 h to accli-
`mate to the surroundings. The treatment of the immature rats began 24 h
`after arrival to ensure that the rats did not reach sexual maturity before
`the completion of treatment. After the acclimation period, the animals
`were treated once a day for 4 days with the compound(s) of interest.
`Doses were prepared based on milligrams per kg mean group BW.
`Administration of the compound was either by sc injection (sc) of 0.2 ml
`in the nape of the neck or intragastrically by gavage (orally) in a volume
`of 0.5 ml. A vehicle control group was included in all experiments.
`Approximately 24 h after the final treatment the animals were killed by
`C02 asphyxiation. After death, the uteri were removed from the animals,
`drained of fluid, stripped of remaining fat and mesentery, and weighed.
`Plasma cholesterol measurements
`
`Blood samples were collected by cardiac puncture after death into
`vacuum tubes containing EDTA to prevent coagulation. The samples
`were centrifuged (1000 X g, 10 min), and the plasma was removed and
`placed in fresh tubes. Total cholesterol was determined in whole plasma
`using the Boehringer Mannheim Cholesterol/ HP system pack (Boehr-
`inger Mannheim Diagnostic Laboratory Systems, Indianapolis, IN) and
`the Boehringer Mannheirn Hitachi 911 Analyzer (Boehringer Mannheim
`Diagnostic Laboratory Systems) by the Cardiovascular Division, Wyeth-
`Ayerst Research (Princeton, NJ). HDL was determined in plasma from
`which the LDL and very low density lipoprotein were precipitated using
`the phosphotungstic acid /magnesium chloride precipitation method
`with the HDL-Cholesterol system pack as describedby the manufacturer
`(Boehringer Mannheim Diagnostic Laboratory Systems). Briefly, 200 yd
`plasma were mixed with 500 Ml precipitation reagent. The samples were
`incubated at room temperature for 10-30 min, then centrifuged at 2000 X
`g for 10 min. The supernatant solutions were removed and analyzed for
`cholesterol as described above for total cholesterol. The Boehringer
`Mannheim reagent composition for cholesterol measurement is identical
`in both kits. The kits were validated for cholesterol measurement using
`rat serum, with an intraassay coefficient of variation of l .l 0/0 and an
`interassay coefficient of variation of 1.8%. The reportable range for total
`cholesterol is 3-800 mg/ dl, and that for HDL cholesterol is 3-150 mg/dl.
`
`Statistical analysis
`The data for uterine wet weights and plasma cholesterol levels were
`heterogeneous between the doses. Therefore, the uterine weights were
`
`transformed by logarithms, and cholesterol levels were transformed by
`square root to stabilize the variability. After transformation, the Huber
`M-estimation weighting was used to down-weight the outlying trans-
`formed observations (24). IMP software (SAS Institute, Cary, NC) was
`used to analyze the transformed and weighted data for both the one-way
`ANOVA and the nonlinear dose-response curves. In all cases, the dose-
`response curves were nonlinear; that is, when the response was plotted
`against the log of the concentration, the curves were sigmoidal. Dose-
`response data are calculated and expressed as the ECSC (mean : sE) for
`uterotropic effects and the IC50 (mean : 51:1) for lipid—lowering effects.
`The EC50 and IC50 values were calculated using the four-paialrieter
`logistic model that calculates the minimum, maximum, I-Iill’s coefficient,
`and ED50 (25). In cases where the dose-response curves did not plateau
`or the response was too shallow, the program was unable to calculate
`an EC50 or IC5O value. In these cases, the EC50 or IC50 values were
`estirnated graphically.
`
`Results
`
`Plasma lipid and uterotropic effects of EE, 17a-E2, and
`1 7B-E2
`
`The effect of EE, administered either orally or sc to adult
`ovariectomized rats, is shown in Fig. 1. When administered
`orally, uterine wet weight increased in a dose—dependent
`nanner (Fig. 1A), and both plasma total and HDL cholesterol
`evels decreased similarly (Fig. 1, B and C). The mean utero-
`ropic EC50 for four separate experiments was 100.8 [Lg / kg
`3W, with IC50 values of 21.1 and 17.7 pig/kg BW for total and
`-IDL cholesterol, respectively. When EE was administered
`via the sc route (five separate experiments), the mean EC50
`"or uterine wet weight increase over vehicle was 0.3 ug/ kg
`3W, 300-fold lower than when E3 was administered by ga-
`vage (Table 1). However, the IC50 values of E3 for plasma
`otal and DL cholesterol lowering were the about the same
`as when ,1: was administered orally (21.6 and 15.1 ug/kg
`3W, respectively). The data for HDL are shown 1ere,but will
`rot be shown for subsequent experiments because in all cases
`he effect o: estrogens on plasma HDL was similar to that on
`plasma total cholesterol.
`To determine whether this was a unique property of E3,
`we evaluated l7(r—E2 and l7B—E2 in a similar study. The two
`compounds were administered at doses of 0.01, 0.5, and 5.0
`mg / kg BW, both orally and sc. As with EE, the effects of both
`compounds were more potent on the uterus when they were
`administered sc, yet their potencies for lipid lowering were
`the same regardless of the route of administration (Fig. 2).
`The estrogens 17oz—E2 and 17/3—E2 were also run in full
`dose-response curves using the sc route of administration.
`The EC50 values for uterine wet weight increase over vehicle
`were 207 and 0.67 rig/kg BW, respectively (Table 1). This
`difference in potencies of 1701-E2 and 170-E2 in the uterus
`correlates well with their relative affinities for the ER. How-
`
`ever, the potencies of the two compounds in the liver, as
`assessed by plasma total cholesterol levels, were only 2-fold
`different; IC50 values were 1414 and 665 p.g/ kg BW for 17a-E2
`and 17,B—E2, respectively (Table 1).
`
`Regulation of lipid levels in immature rats
`
`To extend our studies to the immature rat model, we ran
`dose-response curves for 17oz-EE in 19—day—old rats. The com-
`pound was administered by gavage a doses ranging from
`1-5000 ;rg/ kg BW. As expected, 17rr-EE increased uterine
`wet weight with an EC50 of 8 mg/ kg BW (Fig. 3). I-lowever,
`
`Astrazeneca Ex. 2110 p. 2
`
`
`
`1554
`
`ESTROGEN EFFECTS ON PLASMA CHOLESTEROL LEVELS
`
`Endo o 1997
`Vol 138 I No 4
`
`A
`
`400
`
`levels at the higher dose (Table 2). Dexamethasone signifi-
`cantly decreased body weight about 10% and 30% at 0.05 and
`5.0 mg/kg, respectively, but had no effect on uterine wet
`weight. Dexamethasone also significantly (P S 0.05)
`in-
`creased LDL cholesterol levels at 5.0 mg/kg BW (Table 2).
`Progesterone had no effect on uterine wet weight or plasma
`cholesterol levels (Table 2).
`
`Plasma lipid and uterotropic effects of antiestrogens
`
`The estrogen antagonists tamoxifen, raloxifene, and ICI
`182,780 were also evaluated in this model. Tamoxifen was a
`partial agonist in the uterus when administered sc (Fig. 4A).
`However, its efficacy was only about 20% that of 17oz-EE. It
`had limited ability to lower plasma cholesterol levels; treated
`levels differed significantly from the control Values only at
`1.0 and 10.0 mg/kg 3W. Although the decrease in plasma
`cholesterol was signficant, it was small compared to the
`decrease evoked by he estrogens examined. Total choles-
`terol levels dropped from the control level of 82 mg / dl to 65
`and 52 mg/dl at 1.0 and 10.0 mg/kg BW, respectively (Fig.
`4B). Tamoxifen is me abolized in the liver to its active form,
`4—hyd oxytamoxifen 26). Therefore, we ran a dose—response
`curve with tamoxifen administered by gavage. Unlike
`17oz-E3 and the other compounds, the route of administra-
`tion d "d not affect the potency of the compound in either the
`uterus or liver (Fig. 1).
`Raloxifene, administered sc, also lowered plasma cholesterol
`at all c oses tested (0.0 5, 0.05, 0.5, and 5.0 mg/ kg BW). How-
`ever, tie reduction was small (Fig. 5), lowering total cholesterol
`from the control level of 95 to 64 mg/dl at the 5.0 mg/ kg BW
`dose. Raloxifene also caused a small, but significant, increase in
`uterine wet weight at 0.5 and 5.0 mg/ kg BVV (Fig. 5).
`The pure antiestrogen lCl 182,780 was tested first for poten-
`ial estrogen agonist activity. Unlike the other antiestrogens
`ested, [Cl 182,780 alone had no effect on either uterine wet
`weight or plasma cholesterol even at 5.0 mg/kg BW (Fig. 6, A
`and B). lCl 182,780 was also run as an antagonist against 17a-EE.
`n this experiment 17oz-EE was administered by gavage at 0.1
`ng/ kg BW. This dose,when administered orally, was about the
`ECSO dose for uterine wet weight increase and the lC80 dose for
`ipid lowering. ICI 182,780 was administered sc at doses rang-
`ing from 0.05—5.0 mg/kg BW. When the two compounds were
`coadministe‘ed, lCl 182,780 blocked the uterine wet weight
`'ncrease induced by SE (Fig. 7A). It also blocked the lipid-
`owering effect of E3 (Fig. 7B), suggesting that E3 is acting
`hrough the ER to lower plasma cholesterol levels. The blockage
`of lipid lowering was maximal at about 1.0 mg/kg; however,
`‘t was not complete even at the 5.0 mg/ kg dose.
`
`Discussion
`
`Although the effects of estrogens in the liver and, in par-
`ticular, the involvement of estrogen in reducing plasma LDL
`cholesterol levels have been known for many years, the
`mechanism by which estrogens reduce LDL cholesterol is not
`well defined, especially at the molecular level. Studies using
`high doses of estrogens have indicated that the up—regulation
`of hepatic LDL receptors is the primary mechanism respon-
`sible for the lipid-lowering effect (13, 14). With the present
`studies we have attempted to better characterize the rat as a
`
`Astrazeneca Ex. 2110 p. 3
`
`
`
`300
`
`200
`
`100
`
`o
`
`o
`
`10-4
`
`1o*3
`
`10-2
`
`10-1
`
`100
`
`10‘
`
`0
`
`104
`
`1o*3
`
`10‘?
`
`104
`
`10°
`
`101
`
` o
`
`10-4
`
`10-3
`
`10-2
`
`10-1
`
`10°
`
`101
`
`17oz-Ethinyl estradiol (mg/kg BW)
`
`FIG. 1. Dose-response curves for EE on uterine Wet weight (A) and
`total (B) and HDL (C) cholesterol. EE was administered either in-
`tragastrically by gavage (
`) or so injection ( 9 :1 in 10% ethanol in corn
`oil vehicle once a day for4 days.P0i7Lts are the mean from four animals
`per group shown with the SE.
`
`unlike the adult rat, total and HDL cholesterol were un-
`changed, even at the 5.0 mg/ kg dose (Fig. 3).
`
`Steroid specificity
`
`The steroid specificity of the lipid-lowering effect was also
`examined. The animals were treated with testosterone pro-
`pionate, dexamethasone, and progesterone by sc adminis-
`tration at doses of 0.05 and 5.0 mg/ kg BW. Testosterone
`propionate significantly increased uterine wet weight at 5.0
`mg/kg BW and had a marginal effect on plasma cholesterol
`
`E’
`2»(D
`E
`
`E.
`
`E3 E
`
`ED
`
`B
`
`E2(I)
`2O_C
`<6
`*5I-
`
`0"
`
`C
`
`E2U
`
`?EO
`
`.C
`
`O_
`
`lOI
`
`
`
`ESTROGEN EFFECTS ON PLASMA CHOLESTEROL LEVELS
`
`1555
`
`TABLE 1. Summary of EC50 and lC50 values for 17a-ethinyl estradiol, 173-estradiol, and 1701-estradiol
`
`Compound
`1701-Ethinyl estradiol
`
`1713-Estradiol
`17a-Estradiol
`
`Route
`Oral“
`scb
`sc
`sc
`
`EC50 and TCEO values : SE are shown.
`“ Mean and SE of four separate experiments.
`1’ Mean and SF) of five separate experiments.
`
`EC50 uterine wt
`(Mg/kg BW)
`100.8 : 20.6
`0.30 : 0.15
`0.67 : 0.087
`207 : 37.2
`
`ICED total cholesterol
`(Mg/kg BW)
`21.1 : 6.4
`21.6 : 7.8
`665 : 43.8
`1414 : 146
`
`A
`
`400
`
`3
`E
`- 300
`I-5,
`0’
`3 200
`6
`3
`Q)
`E 100
`E
`3
`
`0
`
`.
`.
`Uterine Wet Weight
`
`Total Cholesterol
`
`100
`
`
`
`-_~
`E
`°’ 75
`E,
`3
`B 50
`3
`3
`.C
`0 25
`E
`o
`" o
`
`0
`
`1o-2
`
`10-1
`
`100
`
`101
`
`0
`
`1o-2
`
`10-1
`
`100
`
`101
`
`1713-estradiol (mg/kg BW)
`
`17g_eSt|-adiol (mg/kg BW)
`
`B
`
`400
`
`Uterine Wet Weight
`
`100
`
`3
`E
`S 300
`0’
`E 200
`as
`3
`E 100
`E
`D
`
`2
`
`»_~
`E
`E 75
`6
`S3 50
`$
`6
`5 25
`E
`
`o
`
`10-2
`
`10-1
`
`10°
`
`101
`
`Total Cholesterol
`
`
`
`170:-estradiol (mg/kg B\N)
`
`170i-estradiol (mg/kg BW)
`
`FIG. 2. The effect ofthe route of administration of 17B-E2 (A) and 1701-E2 (B) on uterine wet weight increase and plasma total cholesterol levels.
`The compounds were administered either intragastrically by gavage (9) or by sc injection (
`) in 10% ethanol in corn oil Vehicle once a day for
`4 days. Points are the mean from six animals per group for the 17[3-E2-treated animals and seven animals per group for the 17oi-E2-treated
`animals, shown with the SE.
`
`model system for the lipid-lowering effects of estrogens, to
`further define the mechanism of action, and to address the
`role of the hepatic ER in this response.
`The rat has noted differences and shortcomings as a model
`for human cholesterol metabolism that must be considered
`
`when using the rodent model. Most notably, in the rat l-lDL
`is the predominant form of cholesterol in plasma, comprising
`about 60-70% of the total cholesterol pool. Moreover, both
`LDL and HDL cholesterol levels decrease after estrogen
`treatment; in humans, estrogens decrease plasma LDL, but
`increase plasma HDL (7-9). One mechanism that may ex-
`plain the difference in HDL metabolism is that rat HDL
`contains higher amounts of apoprotein E than does human
`
`dDL (10). The rat LDL receptor has high affinity for apo-
`protein E. Therefore, l-lDL particles containing apoprotein E
`are cleared from the blood at a higher rate in rats than in
`iumans after estrogen treatment (10). A second mechanism
`hat may contribute to the decrease in plasma HDL in rats is
`he effect of estrogen on the enzymes involved in l-lDL me-
`abolism. It has been reported that estrogens decrease li-
`poprotein lipase (LPL) activity in rats (27, 28). Decreasing
`EL activity lowers plasma HDL levels. Moreover, hepatic
`ipase, which is down—regulated after estrogen treatment in
`uunans (29), is not regulated by estrogen in rats (30). There-
`rore, clearly, the effects of estrogen on HDL metabolism
`cannot be addressed in this model. Even with the differences
`
`Astrazeneca Ex. 2110 p. 4
`
`
`
`1556
`
`ESTROGEN EFFECTS ON PLASMA CHOLESTEROL LEVELS
`
`Endo o 1997
`Vol 138 I No 4
`
`80
`
`(mg)
`UterineWetWeight
`.9
`
`25
`
`0 E
`
`B
`
`100
`
`SU)
`E 75
`I3(D
`G
`EO.C
`
`50
`
`o
`
`10-4
`
`10-3
`
`10-2
`
`10-1
`
`100
`
`10‘
`
`Tamoxifen citrate (mg/kg BW)
`FIG. 4. The effect ofthe antie strogen tamoxifen citrate on uterine wet
`weight (A) and plasma otal cholesterol (B). Tamoxifen citrate was
`administered either sc (
`) or intragastrically by gavage (Q) in 10%
`ethanol in corn oil for 4 days. Points represent the mean from eight
`animals per group, shown with the SE. *, Significantly different from
`vehicle control (P < 0.05) with oral administration. T, Significantly
`different from vehicle control (P < 0.05) with so administration.
`
`was required for the effects of the estrogens, the potencies
`would differ when the compounds were administered orally
`or sc. Therefore, in the rat, the cholesterol-lowering effect of
`estrogen does not require the first pass through the liver.
`Our studies demonstrate that estrogens have no effect on
`plasma cholesterol levels in the immature rat. To our knowl-
`edge, this is the first report of this finding. It has previously
`oeen shown that there is developmental regulation of com-
`oonents of the plasma lipoprotein particles in rats. The mes-
`senger RNA levels for both apoprotein Al and All rapidly
`change between days 20 and 40, the period when the animals
`go through sexual maturation (35). Also, platelet—activating
`factor-acetylhydrolase, an enzyme that is associated with
`DL and HDL particles, is estrogen regulated in adult rats,
`out not in immature rats (36). lt has been reported that ER
`evels in the liver are developmentally regulated (37), low in
`'mmature animals and higher in adult animals, and may
`account for the developmental regulation. We are interested
`'n this developmental regulation and are continuing to pur-
`sue its mechanism.
`
`The effects of both tamoxifen and raloxifene on plasma lipid
`evels were less than reported previously (11, 18-20). This is
`
`Astrazeneca Ex. 2110 p. 5
`
`
`
`UterineWetWeight(mg)<—-—)
`
`60
`
`(mg/dl]
`PlasmaCholesterol
`
`
`40
`
`20
`
`(-4:-—)
`
`17oc-EE (mg/kg BW)
`FIG. 3. The effect of EE on uterine wet weight (I) and plasma total
`cholesterol levels (
`) in immature Sprague-Dawley rats. Nineteen-
`day-old Sprague-Dawley rats were treated for 4 days with EE in 10%
`ethanol in corn oil intragastrically by gavage. Points are the mean
`from eight animals per group, shown with the SE. Only the uterine
`weights are significantly different from the vehicle control value (*,
`P < 0.05).
`
`TABLE 2. Steroid specificity of lipid lowering in adult
`ovariectomized Sprague-Dawley rats
`
`Com ound
`p
`Vehicle
`Testosterone
`propionate
`
`Dexamethasone
`
`Progesterone
`
`Dose
`(mg/kg BW)
`0.0
`0.05
`
`5.0
`0.05
`5.0
`
`0.05
`5.0
`
`Uterine Wet wt
`(mg)
`85.9 1 7.2"
`86.3 : 4.4
`
`246.3 i 16.3“
`90.5 : 2.1
`78.9 : 1.7
`
`95.3 : 5.2
`100.9 : 4.1
`
`Plasma total
`cholesterol (mg/dl)
`80.3 : 3.2
`77.6 : 8.5
`
`681 i 2.4“
`86.6 : 5.0
`117.8 : 6.2“
`
`853 : 3.2
`82.3 : 3.8
`
`Compounds were administered sc. Shown are the mean : SE for
`groups of eight animals.
`“ Significantly different from vehicle control (P < 0.05).
`
`in HDL metabolism, the rat system provides a good model
`to study the mechanism of LDL lowering. There is evidence
`that at least some aspects of the mechanism of LDL lowering
`are similar in rats and humans. The LDL receptor is up-
`regulated in rats; there is evidence for similar regulation in
`the human hepatoma cell line HepG2 and in human liver
`homogenates (13—16). The validity of the model is further
`supported by the fact that compounds that reduce plasma
`total cholesterol in the rat model, such as EE, 17B-E2, tamox-
`ifen, and raloxifene, have beneficial effects on plasma cho-
`lesterol profiles when administered to humans (2, 22, 31).
`There are conflicting reports as to whether, in humans, the
`beneficial effects of estrogen on plasma cholesterol requires
`the ”first pass" through the liver. There are reports demon-
`strating that when estrogens are administered through trans-
`dermal patches, the compounds have little effect on plasma
`cholesterol levels (32). Other reports demonstrate significant
`effects of estrogens on plasma cholesterol when adminis-
`tered by either an oral or a transdermal route (33, 34). In the
`rat, our studies demonstrate that the potencies of five dif-
`ferent estrogens on cholesterol lowering are unaffected by
`the route of administration, lf the first pass through the liver
`
`
`
`
`
`150
`
`100
`
`3,
`E
`5 R100
`%
`09
`E
`E
`E
`5
`:>
`
`o
`
`so
`
`
`
`010-3
`
`10-2
`
`10-1
`
`100
`
`101
`
`50
`
`s
`75 E’
`1’
`g T
`E
`-'4
`0
`|
`25 5 -
`65
`E(D
`o E
`
`Raloxifene (mg/kg BW)
`FIG. 5. The effect of raloxifene on uterine wet weight (I) and plasma
`total cholesterol levels (
`). Raloxifene was administered by sc injec-
`tion in 10% ethanol in corn oil for 4 days. Points represent the mean
`from eight animals per group, shown with the SE. *, Uterine wet
`Weight significantly different fro111 that in the vehicle control (P <
`0.05). +, Total cholesterol significantly different from the vehicle con-
`trol value (P < 0.05).
`
`probably due to the fact that the duration of our treatment was
`only 4 days, much shorter than in previous studies. cz's-tamox-
`ifen, when administered at 0.5 mg/ kg BW for 12 days, de-
`creased total plasma cholesterol 65% (19). Similarly, when ta-
`moxifen citrate was administered weekly at 20.0 mg/kg BW for
`4 weeks,both total and HDL cholesterol levels decreased about
`50% from control levels (18). Raloxifene has also been reported
`to lower total cholesterol in rats when administered daily for 5
`weeks (11, 20). Clearly, the shorter treatment time produced a
`much smaller response than the long term treatment. However,
`the 4-day period is long enough to see significant lowering of
`plasma lipids.
`Interestingly, the potencies of 1701-E2 and 173-E2 in the
`liver and uterus are very different. The potencies we have
`seen in the uterus correspond well with the affinities of these
`two ligands for the ER (23). However, the difference in lC50
`values for these compounds for lipid lowering is only 2-fold.
`It is not believed that the liver has the enzymatic capacity to
`isomerize the 17a-isomer to the 17[3-isomer. Therefore, the
`mechanism for estrogenic effects on lipid lowering may be
`different from the mechanisms involved in the uterus.
`To address the issue of whether the classical ER is medi-
`
`ating the lipid-lowering effect of estrogens, we used the pure
`antiestrogen ICI 182,780. This compound is a potent anties—
`trogen with little known agonistic activity and is believed to
`act specifically through the ER (38). ICI 182,780 administered
`alone had no effect on either uterine wet weight or plasma
`cholesterol levels, supporting its profile as a pure antiestro-
`gen. However, when administered along with EE, it was able
`to block the effect of EE on both uterine wet weight and
`plasma cholesterol. However, the lipid levels never return to
`the vehicle-treated control levels when lCl 182,780 is used as
`an antagonist. It is not clear whether this residual response
`represents an effect that is mediated by a nonreceptor mech-
`anism or the biological variability of the system. This is the
`first report of the effect of ICI 182,780 on rat liver and, in
`particular, plasma cholesterol levels. More importantly, it
`provides evidence for the involvement of the ER in control-
`ling plasma lipid levels.
`
`ESTROGEN EFFECTS ON PLASMA CHOLESTEROL LEVELS
`
`1557
`
`A
`
`'3 200
`
`150
`
`100
`
`50
`
`D
`
`E E
`
`)
`G)
`
`EE
`
`E(D
`-E
`E’.D
`
`Veh
`
`Se
`'
`
`0.05
`
`1.0
`0.5
`lC|182,78O
`
`5.0
`
`100
`
`,
`E’(I)
`'27,‘
`2O.C
`O 25
`
`75
`
`50
`
`0
`
`B A
`
`EB E
`
`EO
`l"
`
`Veh
`
`EE
`0.1
`
`0.05
`
`0.5
`
`1.0
`
`5.0
`
`ICI 182,780
`FIG. 6. The effect of ICI 182,780 on uterine wet weight (A) and
`plasma total cholesterol levels (B). ICI was administered so at doses
`of 005-5 .0 mg/kg BW in 10% ethanol in corn oil vehicle once a day for
`4 days (n = 6 animals/group). Bars represent the SE. *, Significantly
`different from vehicle control (P S 0.05).
`
`ln summary, we have further characterized the rat as a
`nodel for estrogen-mediated plasma cholesterol lowering. lt
`10W seems likely that estrogens are functioning through the
`ER, but there are still many questions that need to be ad-
`dressed. Primarily, what are the molecular targets through
`which estrogens regulate plasma cholesterol levels? The LDL
`‘eceptor is one target already identified, but are there others?
`What relevance do the targets in rats have in regulating
`cholesterol levels in humans? Is there a non—'ER—mediated
`
`component involved in the regulation, and wha is the mech-
`anism of the developmental regulation of the estrogen-in-
`duced lipid lowering? Studies are in progress to address
`some of these important questions.
`
`Acknowledgrnents
`We gratefully acknowledge the scientific input and technical exper-
`tise of Dr. Steven Adelman and the Cardiovascular Group of Wyeth-
`Ayerst Research. We also acknowledge the expertise of the Wyeth-
`Ayerst Research Bioresources Department for its excellent animal care
`and technical assistance, and the Wyeth-Ayerst Research Biometrics
`Department for its assistance with statistical analysis.
`
`Astrazeneca Ex. 2110 p. 6
`
`
`
`1558
`
`ESTROGEN EFFECTS ON PLASMA CHOLESTEROL LEVELS
`
`Endo o 1997
`Vol 138 I No 4
`
`lipoproteins: unresolved issues in biology and clinical practice. Ann NY Acad
`Sci 592:272—285
`. Windler EET, Kovanen PT, Chao Y-S, Brown MS, Havel RI, Goldstein IL
`1980 The estradiol—stimulated lipoprotein receptor o rat liver. I Biol Chem
`255:10464—10471
`. Black L], Sato M, Rowlcy ER, Magcc DE, Bckcle A, Williams DC, Cullinan
`GI, Bendele R, Kauffman RF, Bensch WR, Frolik CA, Termine ID, Bryant HU
`1994 Raloxifene (LY139481 HCI) prevents bone loss and reduced serum cho—
`lesterol without causing uterine hypertrophy. I Clin nvest 93:63—69
`. Washburn SA, Adams MR, Clarkson TB, Adelman SI 1993 A con]ugated
`equine estrogen with differential effects on uterine weight and plasma cho—
`lesterol in the rat. Am I Obstet Cynecol 169:251—256
`Kovanen PT, Brown MS, Goldstein IL 1979 Increase binding of low density
`lipoprotein to liver membranes from rats treated wit
`17a—ethinyl estradiol.
`J Biol Chem 254:ll3fi7—1l373
`. Chao Y-S, Windler EE, Hen GC, Havel RI 1979 Hepatic catabolism of rat and
`human lipoproteins in rats treated with 17a—ethinyl estradiol.
`I Biol Chem
`254:11360—11366
`Semenkovich CF, Ostlund RE 1987 Estrogens induce ow—density lipoprotein
`receptor activity and decrease intracellular cholesterol in human hepatoma cell
`line HepG2. Biochemistry 26:4987—4992
`. Nanjee MN, Koritnik DR, Thomas], Miller ME 1990 Iormonal determinants
`of apolipoprotein B, E receptor expression in human liver. Positive association
`of receptor expression with plasma estrone concentration in middle—aged/
`elderly women Biochim Biophys Acta 1046:151—158
`. Ma PTS, Yamamoto T, Goldstein IL, Brown MS 1986 Increased mRNA for
`low density lipoprotein receptor in livers of rabbits treated with 17o¢—ethinyl
`estradiol. Proc ,\latl Acad Sci LPSA 83:792—796
`Gold E, Stapley S, Goulding A 1994 Tamoxifen and norethisterone: effects on
`plasma cholesterol and total body calcium content in the estrogen deficient rat.
`1-lorm Metab Res 26:1UU—103
`Dunati RI, Harper I’V, Hughes A, Hay RV 1990 Serum cholesterol— and
`apolipoprotein B—lowering effects on cis—tamo><ifen. Arteriosclerosis 10:822A
`(Abstract)
`Evans G, Bryant HU, Magee D, Sato M, Turner RT 1994 The effects of
`raloxifene on tibia histomorphometry in ovariectomized rats. Endocrinology
`134:2283—2288
`McDonald CC, Stewart HI 1991 Fatal myocardial infarction in the Scottish
`adjuvant tamoxifen trial. Br Med J 3U3:435—437
`Draper M, Boss SM, Husler WI, Neild IA 1994 Effects of raloxifene HCI on
`serum markers of bone and lipid metabolism—dose response relationship.
`Calcif T