ICI 182,780 Penetrates Brain and Hypothalamic Tissue and Has Functiona…ter Systemic Dosing -- Alfinito et al. 149 (10): 5219 -- Endocrinology
Articles by Alfinito, P. D.
Articles by Deecher, D. C.
Endocrinology, doi:10.1210/en.2008-0532
Endocrinology Vol. 149, No. 10 5219-5226
ICI 182,780 Penetrates Brain and
Hypothalamic Tissue and Has
Functional Effects in the Brain
after Systemic Dosing
Peter D. Alfinito, Xiaohong Chen, James Atherton,
Scott Cosmi and Darlene C. Deecher
Women’s Health and Musculoskeletal Biology (P.D.A., X.C., S.C.,
D.C.D.), Drug Safety and Metabolism (J.A.), Wyeth Research,
Collegeville, Pennsylvania 19426
Address all correspondence and requests for reprints to: Darlene C.
Deecher, Ph.D., Wyeth Research, RN 3164, 500 Arcola Road,
Collegeville, Pennsylvania 19426. E-mail:
deeched@wyeth.com.
Previous reports suggest the antiestrogen ICI 182,780 (ICI) does not
cross the blood-brain barrier (BBB). However, this hypothesis has
never been directly tested. In the present study, we tested whether
ICI crosses the BBB, penetrates into brain and hypothalamic tissues,
and affects known neuroendocrine functions in ovariectomized rats.
ICI crosses the BBB and penetrates into brain and hypothalamic
tissues
An earlier report suggested that ICI did not cross the BBB of OVX
rats because it failed to block nuclear uptake of [3H]estradiol in
hypothalamic tissue after once daily dosing at 1.0 mg/kg sc for 3 d
(2). However, the ability of ICI to cross the BBB and penetrate brain
tissues was not directly tested in these experiments. In the present study, the same dosing
paradigm was followed as reported by Wade et al. (2) (Fig. 1A), and ICI levels were
measured in plasma and brain (total brain minus hypothalamus and pituitary) and hypothalamic
tissues over time. The ICI compound was detected in all samples and at all time points tested
(Fig. 1B). The concentrations and pharmacokinetic profiles of ICI in plasma, brain, and
hypothalamus were found to be similar (Fig. 1B and Table 1 ). ICI levels were stable in
plasma and brain and hypothalamic tissues over the entire 24-h testing period, and the
concentrations of ICI in plasma and hypothalamic tissue were similar at the 24-h time point.
TABLE 1. Pharmacokinetic profile of ICI (1.0 mg/kg, sc, 3 d) in
plasma and brain and hypothalamic tissues of OVX Sprague Dawley
rats over time
The ratio of brain to plasma exposure is an indication of a compound’s ability to cross the
BBB. Based on 24-h exposure values, brain and hypothalamus to plasma ratios for ICI were
0.33 and 0.66, respectively (Table 1 ). These results demonstrate that ICI crosses the BBB, is
present in brain and hypothalamic tissues, and persists at a constant level in plasma and brain
and hypothalamic tissues for up to 24 h after systemic dosing.
ICI blocks estrogenic actions in the MD model of hot flush but showed estrogenic-like
effects when administered alone
The MD model of hot flush is based on measuring naloxone-induced increases in TST in MD
OVX rats (17, 23). Previous work indicates that estrogen’s ability to abate TST elevations in
the MD model of hot flush occurs through its actions in the brain (18). To determine whether
11/6/08 8:20 AM
ICI 182,780 Penetrates Brain and Hypothalamic Tissue and Has Functiona…ter Systemic Dosing -- Alfinito et al. 149 (10): 5219 -- Endocrinology
P
3), these results demonstrate that ICI rapidly crosses the BBB and persists for an extended
period of time in plasma and brain and hypothalamic tissues. Consistent with its ability to
penetrate brain tissues, repeated systemic administration of ICI blocked the effects of EE on
naloxone-induced TST elevations in the MD model of hot flush and on body weight change.
Interestingly, ICI administration alone (1.0 or 3.0 mg/kg·d) demonstrated weak estrogenic-like
activity in these models. We conclude that ICI is a brain-penetrable compound that can exert
functional (antiestrogenic and estrogenic) effects in the CNS, and specifically the
hypothalamus, after systemic dosing.
Previous studies have concluded that ICI does not cross the BBB because at up to 1.0 mg/kg·d,
it failed to block uptake of [3H]estradiol into nuclei of hypothalamic cells in OVX rats (2) and
failed to mimic the effects of OVX on body weight gain and plasma gonadotropin levels in
intact female rats (1). However, it is possible that these previous studies did not use a high
enough dose to observe inhibitory effects on these endpoints. For example, despite the presence
of ICI in brain and hypothalamic tissue after systemic administration of 1.0 mg/kg·d, we found
that this dose of ICI did not inhibit the effect of EE on all functional endpoints. The 1.0
mg/kg·d dose of ICI did partially inhibit the effect of EE on TST increases in the MD model
but did not block EE’s effect on body weight change. This functional selectivity may be
explained by the fact that ICI’s inhibitory effect on different estrogen-mediated brain functions
can vary depending on the endpoint being studied. For example, Steyn et al. (28) have shown
that intracerebroventricular administration of ICI inhibited estrogen-induced GnRH pulse
frequency but did not block estrogenic effects on progesterone receptor expression in the
hypothalamus or on antepartum prolactin surges. The authors concluded that there might be a
wide range of sensitivities to ICI in the brain that could cause variable results across different
functional endpoints. Thus, it is possible that inhibition of [3H]estradiol uptake into nuclei of
hypothalamic cells or blockade of estrogen’s effect on body weight change may require higher
levels of ICI than other functional endpoints. This idea is supported by our results showing that
ICI treatment at 3.0 mg/kg·d for 8 d did partially block the effect of EE on body weight change.
Because some previous studies (1, 2) only tested ICI at up to 1.0 mg/kg·d, it is unknown if
higher doses would have inhibited [3H]estradiol uptake into nuclei of hypothalamic cells in
OVX rats, or induced body weight gain or increased plasma gonadotropin levels in intact
female rats.
Several lines of evidence suggest that estrogens regulate body weight primarily through central
mechanisms that reduce meal size (15, 16, 29, 30, 31, 32). Lesions of the ventromedial nucleus
of the hypothalamus blocked the effect of systemically administered EB on body weight change
and food intake in OVX rats (29), infusion of estradiol directly into the paraventricular nucleus
or medial preoptic nucleus of the hypothalamus reduced body weight and/or food intake in
OVX rats (30, 31), and direct administration of EB to the hindbrain just above the nucleus
11/6/08 8:20 AM
ICI 182,780 Penetrates Brain and Hypothalamic Tissue and Has Functiona…ter Systemic Dosing -- Alfinito et al. 149 (10): 5219 -- Endocrinology
tractus solitarius reduced food intake in OVX rats (32). However, these results have not been
reproduced in all laboratories (33), and several peripheral feedback mechanisms have been
hypothesized. Therefore, it remains possible that the ability of ICI to block the effect of EE on
body weight change in the present study could occur through peripheral not central
mechanisms. However, direct involvement of peripheral mechanisms in mediating estrogens
effect on body weight is lacking. It has been shown previously that estradiol treatment does not
inhibit feeding by modulating orosensory stimuli (15, 16). Estradiol treatment can inhibit
ghrelin-induced feeding, but this effect does not occur through reduction in meal size as is well
established for estrogens (15). Estradiol does increase the satiating potency of cholecystokinin,
but this effect likely occurs through an estradiol-induced increase in neuronal activity within
the brainstem, not through regulation of signaling in the periphery (15, 16). Finally, leptin
signaling does not appear to directly mediate estrogen’s effect on body weight because estradiol
has reduced body weight and food intake in both leptin-deficient and leptin receptor-deficient
mice (34). Thus, based on current evidence, estrogenic regulation of body weight appears to be
mediated through central mechanisms, and is an appropriate endpoint for predicting whether
ICI crosses the BBB and exerts functional effects in the CNS.
The effect of ICI on body weight has been reported previously in intact cycling female rats and
OVX estrogen-treated rats (2, 24). In these studies body weight changes were unaffected by
daily ICI treatment at either 1 or 1.5 mg/kg·d (higher doses were not tested), and it was
concluded that ICI did not cross the BBB. Our results are consistent with these studies because
the ability of ICI to block EE’s effect on body weight change was not observed at the 1.0
mg/kg·d dose. However, at the higher dose (3.0 mg/kg·d), ICI treatment did block the effect of
EE on body weight change. These data suggest that in previous studies in OVX rats, ICI may
not have been administered at a high enough dose to block estrogenic effects on body weight
regulation.
The effect of ICI alone on body weight change in OVX rats has also been tested previously (2,
24). Results from these studies also suggest that body weight change is unaffected by ICI
treatment at 1.0 or 1.5 mg/kg·d. These data are in contrast to our finding that ICI had weak
estrogenic-like actions on body weight change at both 1.0 and 3.0 mg/kg·d. The discrepancy
between the current work and previous studies is difficult to reconcile. There are several
technical differences such as rat strain and vendor and total treatment length that might account
for the discrepancy. One additional possibility is that the effect of ICI on body weight may vary
depending on initial body weights. In the present work, initial body weights averaged 215 g,
whereas in both other studies, initial body weights were approximately 270–276 g. It is possible
that in heavier rats, ICI may have a greater volume of distribution, increased sequestration into
adipose tissue, and/or increased plasma clearance. These possibilities are supported by studies
showing that increased body mass can alter the pharmacokinetic properties of some drugs (35).
11/6/08 8:20 AM
ICI 182,780 Penetrates Brain and Hypothalamic Tissue and Has Functiona…ter Systemic Dosing -- Alfinito et al. 149 (10): 5219 -- Endocrinology
Therefore, the net result of these effects could be to reduce ICI exposure in brain tissue and
limit its ability to stimulate ER signaling. Importantly, it is unlikely that rats in our study
inadvertently received EE administration because uterine weights were similar to those from
vehicle-treated rats (Fig. 4 ).
Our results suggest that ICI has relatively low clearance in plasma and brain and hypothalamic
tissues of OVX rats when administered at 1 mg/kg·d for 3 d, and can persist in all three
compartments for at least 24 h after the last dose. In fact, the concentrations of ICI for each
tissue were found to be similar at the 0.5 and 24-h time points (Fig. 1B). This relatively low
clearance suggests that ICI could accumulate, particularly in lipid compartments such as brain
and adipose tissues, after daily systemic administration. Although somewhat speculative, this
type of accumulation could alter the pharmacokinetic and pharmacodynamic properties of ICI
over time. Consistent with this idea, ICI treatment at 1 mg/kg·d for 2 d did not block the effect
of EB on lordosis, ear wiggling, or hops and darts but reduced the effect of EB on all three
parameters when administered at 1 mg/kg·d for 24 d (2). Thus, the potential functional effects
of ICI on CNS-mediated endpoints may depend on both the doses tested and the time period
over which dosing is conducted.
A question that occurs is what are the relative roles of ER and β in mediating the effects of
ICI on the endpoints measured in the current study (i.e. TST regulation in the MD model and
body weight change). Because ICI binds to both receptors with similar affinities (36) and both
receptors appear to have broad distribution in the brain, including the hypothalamus (37), the
relative roles of the and β-subtypes are not easily discerned. Regarding body weight, several
studies suggest that estrogen’s effect is ER mediated. In support of this idea, in two separate
studies, the ER agonist 4,4',4''-(4-propyl-[1H]-pyrazole-1,3,5-triyl)trisphenol decreased food
intake and body weight in OVX rats, whereas the ERβ agonist 2,3-bis(4-hyroxyphenyl)-
propionitrile had no effect (38, 39). In addition, EB had no effect on body weight and food
intake in OVX ER knockout mice, suggesting that the β-subtype is insufficient to mediate the
effect of estrogen on these endpoints (40). Finally, other studies have shown that estrogenic
inhibition of feeding occurs through ER-expressing neurons located in the nucleus tractus
solitarius (32). In contrast to these results, in a single study using oligonucleotide knockdown
of ERs in the brain, only ERβ antisense probes blocked the effect of estradiol on body weight
and food intake (41). However, the ability of their probes to reduce ER expression in the brain
was not reported. Thus, based on current data, it appears that the effect of ICI on body weight
change observed in the current study is mediated through the ER receptor subtype. However,
this hypothesis will need to be confirmed in future studies.
Less is known about the respective roles of the ER and β-receptor subtypes in temperature
regulation. Both receptors have been implicated in the regulation of TST elevations in mice
ICI 182,780 Penetrates Brain and Hypothalamic Tissue and Has Functiona…ter Systemic Dosing -- Alfinito et al. 149 (10): 5219 -- Endocrinology
regulation.
The antiestrogenic properties of ICI in the brain as well as uterine tissues have been well
established (1, 28, 43, 44). However, results from our work and others suggest that ICI may not
be a pure antiestrogen. Intrahippocampal infusion of ICI has mimicked the effect of EB on
place learning in OVX rats (43). In addition, Sibonga et al. (24) have shown that, like 17 β-
estradiol (45), ICI treatment (1.5 mg/kg·d) decreases the cancellous bone formation rate in
OVX rats. Finally, in primary hippocampal neurons, both ICI and 17 β-estradiol promoted
neuronal survival against excitotoxic- and β amyloid-induced cell death, induced rapid calcium
influxes, increased spinophilin and Bcl-2 expression, and increased phosphorylation of ERK2
and Akt (25). Thus, ICI appears to have mixed antagonist and agonist properties, and its
pharmacology now seems to be more similar to other selective ER modulators (SERMs), such
as raloxifene and tamoxifen, then initially reported. The precise mechanisms supporting the
mixed pharmacology of ICI are unknown, however, it may be related to the differential
regulation of ER dimers in the absence or presence of an estrogen. It is well known that ER
dimerization is a key step in the activation of estrogen signaling pathways. Using a yeast two-
hybrid system, Wang et al. (46) found that ICI induced ER dimerization when given alone but
destabilized ER dimers in the presence of an estrogen. Therefore, it is possible that ICI-induced
receptor dimerization could lead to activation of estrogen responsive pathways, whereas
destabilization of ER dimers in the presence of an estrogen would block signaling. In support
of this idea, ICI was found to activate a subset of estrogen-responsive genes in MCF-7 cells, a
breast cancer cell line, grown in hormone-depleted medium (47). Although this explanation
might account for the antagonist and agonist-like effects of ICI observed in the MD model and
on body weight change in the current study, it cannot be broadly applicable to all endpoints
because ICI had no detectable estrogenic-like effect on uterine tissue. Currently, it is not well
understood how the tissue-selective agonist/antagonist properties of SERMs, like ICI,
tamoxifen, and raloxifene, manifest. It has been hypothesized that agonist/antagonist activities
of SERMs result from specific ligand-induced conformational changes in ERs that alter
coactivator/corepressor protein binding, and selectively influence different genomic and/or
nongenomic signaling pathways (48). In addition, cell-specific promoter context could play a
role in determining whether a SERM will elicit estrogenic or antiestrogenic actions.
The clinical use of ICI is not likely to be altered significantly by the results from the present
work. However, our results do offer a mechanistic explanation for the occurrence of hot flushes
in premenopausal women treated with fulvestrant (8). In addition, the use of fulvestrant in
premenopausal women would be expected to induce other CNS-related menopause-like
11/6/08 8:20 AM
ICI 182,780 Penetrates Brain and Hypothalamic Tissue and Has Functiona…ter Systemic Dosing -- Alfinito et al. 149 (10): 5219 -- Endocrinology
Psymptoms such as sleep disturbances, mood changes, loss of energy, weight gain, and
decreased libido (49). Because most of these symptoms are not life threatening, it is unlikely
that they would limit the use of fulvestrant for treating advanced breast cancer patients.
Although we found that ICI has weak estrogenic-like activity on certain CNS-mediated
functions, it is also unlikely that these results will affect the use of fulvestrant for treating breast
cancer patients. Previous work has shown that ICI inhibits human breast cancer cell
proliferation (1), and agonist-like activity would be inconsistent with the studies showing the
utility of fulvestrant in treating ER-positive breast cancer (6, 7).
In summary, we have shown that, in contrast to previous conclusions, ICI is capable of crossing
the BBB, penetrating brain and hypothalamic tissues, and exerting functional effects on
neuroendocrine endpoints after systemic administration. We have also found that ICI is not a
pure antiestrogen, and may have a mix of both agonist and antagonist activities on certain
CNS-mediated functions. Therefore, future studies should consider the potential for ICI to
influence estrogen-related functions in the CNS after systemic dosing.