Directory UMM :Data Elmu:jurnal:B:Biological Psichatry:Vol48.Issue10.2000:

Postmenopausal Women before and after Estrogen

Replacement Therapy

John E. Piletz and Uriel Halbreich

Background: Platelet

a

_2A

-adrenoceptors (

a

_2A

AR) and

imidazoline binding sites (subtype I

1

) have been proposed

as peripheral markers of brain stem receptors that

medi-ate sympathetic outflow and are reported to be elevmedi-ated in

major depression.

Methods: In our study, p[

125

I]-iodoclonidine was used to

assess platelet

a

_2A

AR and I

1

binding sites in healthy

postmenopausal women (n

5

34) compared with healthy

women of reproductive age (n

5

26). Receptor

determi-nations were repeated in 19 postmenopausal women

following 59 – 60 days of estrogen replacement therapy

(ERT; 0.1 mg estradiol transdermal patches).

Results: I

1

binding sites were twofold higher in platelets

of postmenopausal women compared with women of

re-production age but were down-regulated (normalized)

after 59 – 60 days of ERT. All other binding parameters,

including platelet

a

_2A

AR density, were not different

be-tween groups nor were they changed after ERT. Platelet I

1

densities after 59 – 60 days of ERT were positively

corre-lated with plasma luteinizing hormone concentrations.

Conclusions: It is suggested that increased imidazoline

binding sites might be associated with mood and

behav-ioral changes in postmenopausal women. Biol Psychiatry

2000;48:932–939 © 2000 Society of Biological Psychiatry

Key Words: Imidazoline receptors,

a

₂

adrenoceptors,

monoamine oxidase, estrogen, luteinizing hormone,

meno-pause, depression

Introduction

P

ostmenopausal women are at high risk for osteoporosis

(American Medical Association 1984) and

cardiovas-cular disease (Stampfer et al 1991). Accordingly, estrogen

replacement therapy (ERT) has been indicated for the

prevention of osteoporosis and cardiovascular disease in

postmenopausal women. Menopausal status may also

influence cognitive problems (Yaffe et al 1998) and

vulnerability to depression (Halbreich 1997; Halbreich

and Lumley 1993). Some studies have suggested that ERT

might improve cognition (Haskell et al 1997; Sherwin,

1988) and stabilize mood (Ditkoff et al 1991; Klaiber et al

1997) in postmenopausal women.

Clonidine is a centrally active, partial agonist for

a

₂

-adrenoceptors (

a

₂

AR) that elicits a number of

cardio-vascular (Parsons and Morledge 1970), hormonal (Siever

and Uhde 1984), and cognitive effects (Ammassari-Teule

et al 1991). Clonidine’s agonistic activities at

a

₂

AR can be

discriminated from its nonadrenergic activities by

induc-tion of a G-protein “coupled” state of

a

₂

AR, yet not all of

clonidine’s effects are mediated through adrenoceptors

(Piletz et al 1994). Specifically, clonidine has nanomolar

affinity for nonadrenergic I

1

-imidazoline binding sites (I

1

sites: Ernsberger et al 1995a).

The molecular nature and function of I

1

sites has

remained a subject of debate (Ernsberger and Haxhiu

1997) even while attempts to clone I

1

sites are being

reported (Ivanov et al 1998b). In radioligand binding

assays, I

1

sites can be distinguished from three subtypes of

a

₂

AR (

a

_2A

,

a

_2B

, and

a

_2C

AR) by phenylephrine and

guanidine compounds, which, in contrast to all

a

₂

AR

subtypes, lack high affinities for I

1

sites (Ernsberger et al

1995a). Prolonged hypotension is induced by

microinjec-tion of imidazoline compounds, such as clonidine, into the

rostroventral lateral medulla (RVLM) region of the brain

stem (Bousquet et al 1989), the potency of which

corre-lates with affinities of these agents at I

1

sites but not at any

a

₂

AR subtypes (Ernsberger et al 1990). Based on

phar-macologic studies, I

1

sites in the RVLM were proposed to

reside on imidazoline receptors that regulate sympathetic

outflow (Bousquet et al 1989; Ernsberger et al 1990).

More recently, another subtype of I-sites (I

2

sites) has been

identified as a modulatory domain within monoamine

oxidase enzymes (Raddatz et al 1997; Tesson et al 1995).

The latter finding has raised the possibility that

imidazo-From the Department of Psychiatry and Human Behavior, and Departments of Pharmacology and Physiology, University of Mississippi Medical Center, Jackson (JEP) and Biobehavioral Program, Departments of Psychiatry and Gynecology and Obstetrics, State University of New York, Buffalo (UH). Address reprint requests to Uriel Halbreich, M.D., Professor of Psychiatry,

Research Professor of Gyn/OB, Director of Biobehavioral Research, SUNY Clinical Center, Biobehavioral Program, Room BB 170, 462 Grider Street, Buffalo NY 14215.

Received September 23, 1999; revised February 14, 2000; accepted February 15, 2000.

lines might also mediate sympathetic outflow via the

metabolism of catecholamines in the brainstem.

There is also evidence that platelet I

1

and

a

2A

AR

clonidine-binding sites may be codysregulated in

depres-sion. This is based on reports (Piletz et al 1990, 1996a)

that I

1

and

a

2A

AR clonidine-binding sites are elevated in

depressed patients compared with healthy control subjects,

and that both sites are down-regulated following certain

types of antidepressant treatments (Piletz et al 1991,

1996b). Furthermore, platelet

a

_2A

AR and I

1

sites appear to

possess identical binding properties as their counterparts

in the brain (Piletz and Sletten 1993). Elevations in brain

I

1

and

a

2A

AR clonidine-binding sites have also been

suggested (Callado et al 1998; Garcia-Sevilla et al 1996,

1999) in depressed suicide victims compared with

sudden-death matched control subjects. Thus, high densities of I

1

and

a

_2A

AR sites may be associated with depressed mood

(Garcia-Sevilla et al 1999).

Alpha

2

AR in uteri and brain have been reported to be

influenced by gonadal hormones and to regularly fluctuate

along the menstrual cycle (Bottari et al 1983; Orensanz et

al 1982; Roberts et al 1979). On the other hand, we have

found irregular fluctuations of platelet

a

_2A

AR and I

1

sites

during the menstrual cycle in correlation with changes in

plasma epinephrine and norepinephrine levels (Piletz et al.

1998). Women with dysphoric premenstrual symptoms

also exhibited higher [

3

H]-para-aminoclonidine binding to

platelet I

1

and

a

2A

AR sites than control subjects, which

was especially pronounced during the symptomatic late

luteal phase of the disorder (Halbreich et al 1993).

To our knowledge, platelet

a

_2A

AR and I

1

sites have not

previously been studied in postmenopausal women, and

the effects of ERT on these sites have not been reported.

Considering the possible role of these systems in the

modulation of blood pressure and mood, their influence by

menopause and ERT is of interest from both a clinical and

a heuristic perspective.

Methods and Materials

Subjects

Thirty-four postmenopausal women (aged 52.663.4 SD; range 45–59 years) and 26 women of reproductive age (aged 36.16

6.5, range 22– 45 years) participated in the baseline study. Of the postmenopausal women, 19 were retested for radioligand binding on days 59 – 60 of ERT (aged 52.263.4 years). Six postmeno-pausal women completed ERT but were unable to be rescheduled for blood drawing until 2–3 days after estrogen (E2)

discontin-uation (their posttreatment data will not be included here). Nine postmenopausal women discontinued ERT prematurely or were lost to follow-up.

All women of reproductive age as well as postmenopausal women were recruited, evaluated, and treated, and their samples

were drawn and processed at the same site (Biobehavioral Program, State University of New York at Buffalo).

Eligibility of subjects was determined according to several criteria. Postmenopausal women had not experienced menstrual cycles for at least 2 years. They were at least 1 year after cessation of menopausal symptoms (including hot flashes, night sweats, and irritation by vaginal dryness) but were not older than 60 years of age. Luteinizing hormone (LH) and follicle-stimu-lating hormone (FSH) concentrations were within range for menopause, and no plasma progesterone was detected on at least two occasions before the study began. These postmenopausal women had no medical contraindications to ERT. The subjects weighed within 20% of ideal body weight, were physically healthy (including normal blood pressure and pulse), and met research diagnostic criteria (Spitzer et al 1978) for “not currently mentally ill” for at least 2 years before the study. Ten of the enrolled women had lifetime histories of major depressive disorder (MDD) and were therefore treated as a subgroup. Women were excluded if they had any active physical illnesses or other disorders necessitating chemical or somatic treatments or had any Axis I diagnosis of the DSM-III-R (APA 1987) at the time of testing. Postmenopausal subjects were compared with women of reproductive status who had regular menstrual cycles and otherwise met the same inclusion and exclusion criteria. Women of reproductive age were also administered the Premen-strual Assessment Form (Halbreich et al 1982) and daily rating forms (Endicott et al 1986) to exclude subjects with dysphoric symptoms during their midfollicular phases or those with pre-menstrual syndrome. This study was approved by an institutional review committee. All subjects signed consent forms before the study.

slowly through an 18-gauge needle into a heparinized syringe (19 units/mL blood). An additional 5 mL of blood was collected to obtain plasma (stored at270°C) for hormone determinations. The laboratory personnel were unaware of the subject’s age, diagnosis, or treatment (single blind).

Platelet Preparations

The preparation of intact platelets began according to a slight modification of the method of Corash (1980). Platelets were then washed four times in sterile Ca11_/Mg11_{-free Hank’s balanced}

salt solution, pelleted, resuspended in the same solution with 0.1 mmol/L phenyl methylsulfonylflouride, a protease inhibitor, and snap frozen at280°C. After thawing, sonication, and centrifu-gation at 4°C (Piletz et al 1990), the platelet particulates were resuspended in ice-cold reaction buffer (designated washed lysates; formula of reaction buffer given below). Washed lysates were then layered over a discontinuous 14.5% and 34% sucrose gradient and centrifuged at 105,000 g for 90 min at 4°C using a Beckman (Allendale, NJ) SW-28 rotor in a Sorvall OTD60B ultracentrifuge (Dupont, Wilmington, DE). The interface layer, containing the plasma membranes, was diluted in 40 mL of ice-cold water, pelleted, and resuspended in reaction buffer (Piletz et al 1990). The concentration of protein was determined using a Lowry-Biuret reagent kit from Sigma Chemical Co. (St. Louis). Samples were stored at280°C until use.

Binding Assays

Radioligand binding with p[125

I]-iodoclonidine ([125

I]PIC; New England Nuclear, Boston) was according to our previous proce-dure (Ernsberger et al 1995b; Piletz et al 1996b). Specifica2AAR

binding was determined from “total binding” minus “nonadren-ergic binding,” as displaced in the presence of 10 mmol/L norepinephrine. With another aliquot, I1 binding sites were

measured under a mask of 10mmol/L NE (Piletz et al 1996b). Specific I1 binding was then determined from “nonadrenergic

binding” minus “total nonspecific binding,” by displacement in the presence of additional 100 mmol/L moxonidine (kindly contributed by Dr. Siegfried Schafer, Solvay Pharma, Hanover, Germany). In most cases, the binding assays entailed seven concentrations of radioligand per site (Piletz et al 1996b); however, in some cases, a low protein yield permitted only 4 – 6 concentrations of radioligand per site. All measurements were made in hextuplicate (i.e., three total and three nonspecific values for each concentration). Reactions were initiated by adding 0.015– 0.03 mg plasma membrane proteins (up to 0.25 mL) to the radioligand and incubating at 21°C. The reaction buffer was 5 mmol/L HEPES, 0.5 mmol/L MgCl2, 0.5 mmol/L EGTA

(ad-justed to pH 7.4 with 0.1 mmol/L NaOH); with 0.1 mmol/L ascorbic acid added on the day of the binding assay. Trapped plasma membranes were washed three times with ice-cold wash buffer on thick glass fiber filters (Schleicher & Schuell #32, Keene, NH), and radioactivity was counted for 10 min per sample in a Cobra Gamma Counter (Packard Instruments, Meriden, CT) at 80% efficiency. Individual Bmaxand KDvalues

were quantified using LIGAND (McPherson 1985). Data were fit to both one-site and two-site models, but a one-site model was always preferred by an F test.

Hormonal concentrations were determined by radioimmuno-assays (RIA) with commercially available kits. 17-Beta estradiol (E2) was assayed (Diagnostic Products, Los Angeles) with 100ml

aliquots of plasma that were incubated at room temperature for 3 hours in an antibody-coated tube with125

I-free estradiol. Bound estradiol was counted after decantation of the supernatant. The intra- and interassay coefficients of variation (CV) ranged from 4.0% to 8.1%; the sensitivity limit was 8 pg/mL. Plasma progesterone determinations (Diagnostic Products) followed pro-cedures similar to those used for estradiol. The intra- and inter-assay CV ranged from 6% to 10%; the sensitivity limit was 0.05 ng/ml. Plasma testosterone was assayed by a solid phase RIA (Diagnostic Products) in an aliquot of 50mL. Interassay CV ranged from 9%–13%, and intraassay CV ranged from 4% to 6%. The sensitivity limit for the testosterone assay was 0.04 ng/mL. Plasma FSH and LH were also determined with RIA reagents purchased from Diagnostic Products in 200-mL aliquots. The interassay CV ranged from 4% to 6% (FSH) and 2% to 7% (LH), and the intraassay CV ranged from 2% to 4% (FSH) and 1% to 2% (LH). The sensitivity limit for FSH was 0.06 mIU/mL and for LH was 0.1 mIU/mL.

Statistical Analyses

Subject groups were compared with each other using nonpaired Student’s t tests. Pre- to posttreatment values were compared with paired t tests. Pearson’s regression tests were performed for associations of radioligand binding values with hormone values. All p values are two tailed. The data are expressed as means6

standard deviations.

Results

As shown in Table 1 and Figure 1, platelet I

1

density was

significantly higher (pre-ERT) in postmenopausal women

(n

5

34) compared with women of reproductive status

(n

5

26; I

1

B

max

t

5

5.4, p

,

.0001). None of the

other platelet radioligand binding parameters were

signif-icantly different between postmenopausal women and

women of reproductive status (

a

_2A

AR B

max

t

5

0.38 p

5

.7;

a

_2A

AR K

D

t

5

0.68, p

5

0.5; I

1

site K

D

t

5

1.17, p

5

.25). There was no correlation between the

ages of postmenopausal women and any K

D

or B

max

parameters (pretreatment rs:

5

.04 to .28, posttreatment

rs

5 2

.02 to

2

.24).

Before ERT, postmenopausal women had plasma E

2

concentrations of 7.4

6

6.8 pg/mL, progesterone

concen-trations of 0.18

6

0.13 ng/mL, LH concentrations of

81.4

6

10.0 mIU/mL, FSH concentrations of 102.6

6

52.1

mIU/mL, and testosterone concentrations of 0.24

6

0.12

pg/mL. After 59 – 60 days of estrogen treatment, plasma

E

2

concentrations were 63.6

6

39.7 pg/mL (p

,

.0001

vs. baseline), plasma LH concentrations were 9.3

6

5.7

mIU/mL (p

,

.0001 vs. baseline), plasma FSH

concen-trations were 39.2

6

11.3 mIU/mL (p

,

.0001 vs.

baseline), and plasma testosterone concentrations were

0.31

6

0.40 pg/mL (ns). Progesterone concentrations after

ERT were below the sensitivity of the RIA for most

women. Baseline and posttreatment concentrations of E

2

or other hormones did not correlate with each other (r

5

.03, p

5

.86 for E

2

). Plasma E

2

concentrations after 30

days of ERT (61.6

6

36.0 pg/mL) did not differ from E

2

concentrations after 59 – 60 days of ERT (t

5

0.42, p

5

.63).

Correlations between Binding Sites and Hormones

At baseline (pre-ERT), none of the platelet receptor

parameters were significantly correlated with any of the

plasma

hormone

concentrations

in

postmenopausal

women (r values ranged from .1 to .45, ns).

Because ERT induced a down-regulation of platelet I

1

sites (Table 1 and Figure 1), it was anticipated that E

2

concentrations at post-ERT might be negatively correlated

with I

1

B

max

values. In fact, the opposite was observed. A

positive correlation was found between posttreatment E

2

concentrations and I

1

B

max

values at the p

5

.04

significance level (r

5

.49, n

5

18); however,

pre-to-post treatment I

1

B

max

change scores (the

magni-tude of decrease) were positively correlated with

posttreat-ment E

2

concentrations (r

5

.46, p

5

.04).

It was also anticipated that plasma LH concentrations,

FSH concentrations, or both might be positively correlated

with platelet I

1

B

max

values at posttreatment. This was

because all three of these measures appeared to decrease

following ERT. In accord with this prediction, plasma

concentrations of LH after 59 – 60 days of ERT were

highly positively correlated with I

1

B

max

values (r

5

.73,

p

5

.001, n

5

17) (Figure 2). Furthermore, the

pre-to-post treatment I

1

B

max

change scores were

nega-tively correlated with posttreatment LH concentrations

(r

5 2

.48, p

5

.04). Thus, higher E

2

and lower LH

concentrations after ERT were associated with higher

Figure 1. Effect of menopause and estrogen therapy (ERT) on plateleta2-adrenergic and imidazoline-1 sites. [

125

I]PIC, p[125

I]-iodoclonidine; prot, protein.

Figure 2. Platelet imidazoline binding site and plasma luteiniz-ing hormone followluteiniz-ing estrogen replacement therapy in post-menopausal women. prot, protein.

Table 1. Platelets Imidazoline anda2AR Binding

Subjects

Baseline Post-ERT

I1Bmax I1KD a2Bmax a2KD I1Bmax I1KD a2Bmax a2KD

All postmenopausal women (n5 34)

177.6689.3a _4.2

61.2 64.6659.8 0.6660.53 — — — —

Subjects who completed 60 days of E2(n5 19)

188.3663.3b _4.1

61.2 76.6664.0 0.7260.59 122.7658.1 4.461.7 80.7653.7 0.6260.37

Reproductive age women (n5 26) 84.6630.5 4.661.6 60.0631.9 0.5560.61 — — — —

Values represent mean6SD. E2, 17-beta estradiol.ap,.0001 vs. women of reproductive age.

change scores (i.e., more down-regulation) of I

1

B

max

values. No other significant correlation was observed

between any of the hormone concentrations and either I

1

or

a

_2A

AR parameters.

In general, there was no association between blood

pressure or pulse rate with any of the platelet I

1

or

a

2A

AR

parameters (rs: .028 –.184). A negative correlation (r

5

2

.4040, p

5

.037) was observed between platelet

a

_2A

AR affinity (K

D

) and pulse rate at baseline, however.

Subgroup with Lifetime History of MDD

Postmenopausal women with lifetime histories of MDD

(n

5

10) tended to have higher I

1

B

max

values than those

who were never mentally ill (n

5

24), but this did not

reach statistical significance (221

6

132 vs. 159

6

69

fmol/mg protein, t

5

1.62, p

5

.12). Similarly, I

1

K

D

values did not differ (4.1

6

1.0 vs. 4.2

6

1.2, t

5

.3, p

5

.77), nor did

a

_2A

AR K

D

values differ (0.42

6

0.40 versus

0.74

6

0.63, t

5

1.39, p

5

.18) in regard to lifetime

histories of MDD. Postmenopausal women with lifetime

histories of MDD differed in their

a

_2A

AR B

max

values at

baseline, however, which were high (73.8

6

61.2 vs.

30.1

6

16.6 fmol/mg protein, t

5

2.83, p

5

.01). The

two groups (with or without lifetime MDD) did not

significantly differ in baseline hormonal concentrations.

Following 59 – 60 days of treatment with Estraderm, the

women with past MDD (n

5

7) did not differ from those

without lifetime histories of MDD on any of the I

1

parameters; however, after E

2

treatment, the LH

concen-trations of women with histories of MDD decreased to

lower concentrations than did the LH of women who were

never mentally ill (3.9

6

2.4 vs. 11.1

6

5.6 ug/L, t

5

3.23, p

5

.008).

Discussion

We report two main findings. The first is a higher

radioligand binding density of I

1

sites in platelets of

postmenopausal women compared with women of

repro-ductive age (Table 1 and Figure 1). The second is

down-regulation of platelet I

1

binding sites after 59 – 60

days of ERT (Table 1 and Figure 1). These findings are

distinguished from

a

_2A

AR sites on platelets, which were

not different in postmenopausal women compared with

women of reproductive age and were not affected by

ERT.

We suggest that the high density of platelet I

1

binding

sites observed in postmenopausal women might relate to

altered hormonal status, rather than to age per se. This is

for the following reasons: 1) In our study, there were no

within-group correlations between ages and I

1

or

a

2A

AR

binding parameters. 2) In our five previous studies

(Hal-breich et al 1993; Piletz et al 1990, 1996a, 1996b, 1998),

with more than 90 subjects of varying diagnoses (male and

female subjects, ages 20 – 65), no statistically significant

associations have been found between ages and densities

of platelet I

1

or

a

2A

AR sites. 3) Even considering only

those reports (Piletz et al 1996a; 1996b, 1998) that used an

identical binding assay and where a sizeable number of

healthy female control subjects were available (n

5

44),

we could find no association between female ages (23– 63

years) and densities of platelet I

1

or

a

2A

AR sites. To our

knowledge, no other reports exist on aging effects on these

platelet sites. Because no previous reports have even

hinted at an effect of age per se on these sites, the best

explanation for the present data seems to be the

postmeno-pausal status of women and its related hormonal state.

I

1

and

a

2

AR sites are known to be differentially

distributed in the brain (DeVos et al 1994; Kamisaki et al

1990). I

1

sites are almost completely absent in the brain

stem nucleus locus coeruleus, which contains a high

density of

a

_2A

AR (Ernsberger et al 1994). The frontal

cortex is another region of high density of

a

_2A

AR, but

where I

1

sites are of relatively low density (DeVos et al

1994; Kamisaki et al 1990). On the other hand, I

1

sites

have been extensively studied in the RVLM brainstem

nucleus (Bricca et al 1989; Ernsberger et al 1990), where

they are associated with the hypotensive action of

cen-trally applied imidazoline compounds. I

1

sites in the

RVLM might serve a physiologic role in the tonic and

reflex control of the sympathetic nervous system

(Erns-berger and Haxhiu 1997). An endogenous clonidine

dis-placement substance has also been characterized (Atlas

1991) and identified to contain agmatine, a decarboxylated

arginine metabolite (Li et al 1994). Thus, agmatine and the

imidazoline binding sites (I

1

and I

2

) might compose a

neurotransmitter-receptor system. Although agmatine is

now realized to be only one of several molecules that

constitute clonidine displacement substance (Piletz et al

1995; Sun et al 1995), agmatine fulfills several criteria for

a neurotransmitter (Reis and Regunathan 1999).

Al-though we did not measure platelet I

2

sites in our study, it

is conceivable that the platelet I

1

site is physically related

to I

2

sites.

Our finding of no effect of ERT on platelet

a

_2A

AR sites

in postmenopausal women is in accord with a previous

study by Best and coworkers (Best et al 1992) in which

yohimbine binding to platelet

a

_2A

AR sites was unchanged

in postmenopausal women given an estradiol implant (100

mg) for 6 weeks. By contrast, another study reported

increased platelet

a

_2A

AR sites in human male volunteers

treated with E

2

(U’Prichard and Snyder 1979), and three

earlier studies of female rabbits reported that estrogen

treatment decreases

a

_2A

AR sites in platelets (Elliott et al

1980; Mishra et al 1985; Roberts et al 1979). In our study,

and in the study by Best and colleagues (Best et al 1992),

estradiol was administered for a much longer time (42– 60

days) than in the human male study (4 days) or the rabbit

studies (6 –24 days). Therefore, the length of E

2

treatment

and species and gender differences might be critical

factors.

We have also reported (Halbreich et al 1993) higher

platelet

a

_2A

AR and I

1

sites in women with dysphoric

premenstrual syndrome compared with healthy control

women. In that study, elevations in binding were most

pronounced during the symptomatic late luteal phase of

the disorder when plasma norepinephrine and adrenaline

concentrations typically peak (Blum et al 1992). The

literature is uniformly negative for regular, cyclic

fluctu-ations of platelet

a

_2A

AR or I

1

sites along the human

menstrual cycle (Jones et al 1983; Piletz et al 1998;

Stowell et al 1988; Sundaresan et al 1985; Theodorou et al

1987). In certain brain regions and in uteri, however,

a

2A

AR sites may be influenced by gonadal hormones and

appear to fluctuate along the estrus cycle (Bottari et al

1983; Orensanz et al 1982).

One potential paradox was that post-ERT I

1

B

max

values

were positively associated with plasma E

2

concentrations.

This would not be expected if E

2

directly down-regulated

I

1

sites; however, plasma E

2

concentrations were found to

be positively associated with the magnitude of change in I

1

B

max

values, which is in line with an E

2

down-regulation

effect. We have previously reported (Piletz et al 1998)

irregular fluctuations of platelet

a

_2A

AR and I

1

sites over

two menstrual cycles in healthy women, which were not

correlated with phase of the cycle but were instead

correlated with plasma concentrations of adrenaline and

norepinephrine. We have also reported (Ivanov et al

1998a) that immunoreactive I-sites on human

megakaryo-blastoma cells are up-regulated after 6 hours of exposure

to norepinephrine in vitro. Similar treatments with

estro-gen in vitro were without effect (unpublished

observa-tions). Thus, the mechanisms of the down-regulation of

platelet I

1

sites in women after ERT (Table 1) is still

unclear. Future studies are needed to address whether

indirect changes in plasma catecholamines might underlie

platelet I

1

changes.

Another interesting observation was that post-ERT I

1

B

max

values were positively correlated with plasma LH

concentrations (Figure 2). Whether these are causally

linked (i.e., LH or the I

1

site exerts a regulatory effect) or

whether LH and I

1

sites are independently influenced by

E

2

is open to speculation. The literature suggests no effect

of clonidine on LH release (Kaufman and Vermeulen

1989). Any effect of LH on I

1

binding sites therefore is

unknown and will require additional studies (in the same

way that the possible effect of LH on mood and behavior

has not been sufficiently investigated).

It is also interesting to note that the abnormal I

1

B

MAX

values previously reported for depressed patients (Piletz et

al 1996b) are comparable in level to those of healthy

postmenopausal women (Table 1). Of course, exact

plate-let I

1

B

MAX

values are notoriously difficult to duplicate

across studies for technical reasons (Ernsberger et al

1995b). Even slight variations in platelet preparation may

be critical (Corash 1980). In fact, all the platelets in our

current study were prepared at a different site (Buffalo)

than in our previous studies with depressed patients

(Cleveland). Nevertheless, the B

MAX

values in our

post-menopausal women (177.6

6

89.3 fmol/mg protein, n

5

34) are surprisingly close to what we previously reported

for a group of middle-aged male and female depressed

patients (161

6

66.3 fmol/mg, n

5

26) (Piletz et al

1996b). I

1

densities in postmenopausal women after ERT

(122.7

6

58.1 fmol/mg, n

5

19) are also surprisingly

close to what we previously reported for healthy

middle-aged male and female subjects (126

6

63.6 fmol/mg, n

5

18) (Piletz et al 1996b). Furthermore, the few

postmeno-pausal women with lifetime histories of MDD had higher

B

MAX

values for

a

2A

AR (statistically significant) and I

1

sites (trend) compared with women without lifetime

his-tories of MDD. If the platelet I

1

site becomes validated as

a biological marker for depression (Piletz et al 1994), the

present findings in nondepressed postmenopausal women,

as well as the down-regulation by estrogen, will have to be

taken into account.

This work was supported in part by Grant Nos. RO1-MH-45901, RO1-MH-49248, and RO1-MH-57601. We greatly appreciate the clinical assistance provided by Judith Vedella, Ph.D., R.N., and Barbara Bernar-dis, B.S., R.N., the statistical analysis provided by Khalid Bibi, Ph.D., and the laboratory assistance of Jianhua Shen, M.D., Guojing Yuan, and Tammy Stefan.

References

American Psychiatric Association (1987): Diagnostic and

Sta-tistical Manual of Mental Disorders, 3rd ed rev. Washington,

DC: American Psychiatric Press.

Ammassari-Teule M, Maho C, Sara SJ (1991): Clonidine re-verses spatial learning deficits and reinstates 0 frequencies in rats with spatial fornix section. Behav Brain Res 45:1– 8. Atlas D (1991): Clonidine-displacing substance (CDS) and its

putative imidazoline receptor. New leads for further diver-gence ofa2-adrenergic receptor activity. Biochem Pharmacol

41:1541–1549.

Best NR, Rees MP, Barlow DH, Cowen PJ (1992): Effect of estradiol implant on noradrenergic function and mood in menopausal subjects. Psychoneuroendocrinology 17:87–93. Blum I, Nessiel L, David A, Graff E, Harsat A, Weissglas L, et

al (1992): Plasma neurotransmitter profile during different phases of the ovulatory cycle. J Clin Endocrinol Metab 75:924 –929.

Bottari SP, Vokaer A, Kaivez E, Lescrainier JP, Vauquelin GP (1983): Differential regulation ofa-adrenergic receptor sub-classes by gonadal steroids in human myometrium. J Clin

Endocrinol Metab 57:937–941.

Bousquet P, Feldman J, Tibirica E, Bricca G, Molines A, Dontenwill M, Belcourt A (1989): New concepts on the central regulation of blood pressure. Alpha2-adrenoceptors

and “imidazoline receptors. Am J Med 87:10S–13S. Bricca G, Dontenwill M, Molines A, Feldman J, Belcourt A,

Bousquet P (1989): The imidazoline preferring receptor: Binding studies in bovine, rat and human brainstem. Eur

J Pharmacol 162:1–9.

Callado LF, Meana JJ, Grijalba B, Pazos A, Sastre M, Garcia-Sevilla JA (1998): Selective increase of a2A-adrenoceptor

agonist binding sites in brains of depressed suicide victims.

J Neurochem 70:1114 –1123.

Chakravorty SG, Halbreich U (1997): The influence of estrogen on monoamine oxidase activity. Psychopharmacol Bull 33: 229 –233.

Corash L (1980): Platelet heterogeneity: Relevance to the use of platelets to study psychiatric disorders. Schizophr Bull 6:254 –258.

DeVos H, Bricca G, DeKeyser J, DeBacker J, Bousquet P, Vauquelin G (1994): Imidazoline receptors, non-adrenergic idazoxan binding sites, and a2-adrenoceptors in the human

central nervous system. Neuroscience 59:589 –598.

Ditkoff EC, Crary WG, Cristo M, Lobo RA (1991): Estrogen improves psychological function in asymptomatic postmeno-pausal women. Obstet Gynecol 78:991–995.

Elliott JM, Peters JR, Grahame-Smith DG (1980): Oestrogen and progesterone change the binding characteristics ofa -adren-ergic and serotonin receptors on rabbit platelets. Eur J

Phar-macol 66:21–30.

Endicott J, Cohen J, Nee J, Fleiss J, Sarantakos S (1981): Hamilton Depression Rating Scale. Extracted from regular and change versions of the Schedule for Affective Disorders and Schizophrenia. Arch Gen Psychiatry 38:98 –103. Endicott J, Nee J, Cohen J, Halbreich U (1986): Premenstrual

changes: Patterns and correlates of daily ratings. J Affect

Disord 10:127–135.

Endicott J, Spitzer RL (1978): A diagnostic interview: the schedule for affective disorders and schizophrenia. Arch Gen

Psychiatry 35:837– 844.

Ernsberger P, Giuliano R, Willette RN, Reis DJ (1990): Role of imidazole receptors in the vasodepressor response to clonidine analogs in the rostral ventrolateral medulla. J

Phar-macol Exp Ther 253:408 – 418.

Ernsberger P, Graves ME, Graff LM, Zakieh N, Nguyen P, Collins LA, et al (1995a): I1-Imidazoline receptors:

Defini-tion, characterizaDefini-tion, distribution and transmembrane signal-ling. Ann N Y Acad Sci 763:22– 42.

Ernsberger P, Haxhiu MA (1997): The I1-imidazoline-binding

site is a functional receptor mediating vasodepression via the ventral medulla. Am J Physiol 42:R1572–R1579.

Ernsberger P, Haxhiu MA, Graff LM, Collins LA, Dreshaj I, Grove DL, et al (1994): A novel mechanism of action for hypertension control: moxonidine as a selective I-1 imidazo-line agonist. Cardiovasc Drugs Ther 8:27– 41.

Ernsberger P, Piletz JE, Graff LM, Graves ME (1995b): Opti-mization of radioligand binding assays for I-1 imidazoline sites. Ann N Y Acad Sci 763:163–168.

Garcia-Sevilla J, Escriba P, Guimon J (1999): Imidazoline receptors and human brain disorders. Ann N Y Acad Sci 881:392– 409.

Garcia-Sevilla JA, Escriba PV, Sastre M, Walzer C, Busquets X, Jaquet G, et al (1996): Immunodetection and quantitation of imidazoline receptor proteins in platelets of patients with major depression and in brains of suicide victims. Arch Gen

Psychiatry 53:803– 810.

Halbreich U (1997): Role of estrogen in postmenopausal depres-sion. Neurology 48:S16 –S19.

Halbreich U, Endicott J, Schacht S, Nee J (1982): The diversity of premenstrual changes as reflected in the Premenstrual Assessment Form. Acta Psychiatr Scand 65:46 – 65. Halbreich U, Lumley LA (1993): The multiple interactional

biological processes that might lead to depression and gender differences in its appearance. J Affect Disord 29:159 –73. Halbreich U, Piletz JE, Carson S, Halaris A, Rojansky N (1993):

Increased imidazoline anda2adrenergic binding in platelets

of women with dysphoric premenstrual syndromes. Biol

Psychiatry 34:676 – 686.

Haskell SG, Richardson ED, Horwitz RI (1997): The effect of estrogen replacement therapy on cognitive function in wom-en: A critical review of the literature. J Clin Epidemiol 50:1249 –1264.

Holsboer F, Benkert O, Demisch L (1983): Changes in MAO activity during estrogen treatment of females with endoge-nous depression. Mod Probl Pharmacopsychiatry 19:321– 326.

Ivanov T, Feng Y, Wang H, Regunathan S, Reis DJ, Chikkala DN, et al (1998a): Imidazoline receptor proteins are regulated in platelet-precursor MEG-01 cells by agonists and antago-nists. J Psychiatr Res 32:65–79.

Ivanov TR, Zhu H, Regunathan S, Reis DJ, Dontenwill M, Vonthron C, et al (1998b): Co-detection by two imidazoline receptor protein antisera of a novel 85 kilodalton protein.

Biochem Pharmacol 55:649 – 655.

Jones SB, Bylund DB, Rieser CA, Shekim W, Byer JA, Carr GW (1983): Alpha2adrenergic receptor binding in human

plate-lets: Alterations during the menstrual cycle. Clin Pharmacol

Ther 34:90 –96.

(1990): Binding of [3

H]p-aminoclonidine to two sites, a2

-adrenoceptors and imidazoline binding sites: Distribution of imidazoline binding sites in rat brain. Brain Res 514:15–21. Kaufman JM, Vermeulen A (1989): Lack of effect of the

a-adrenergic agonist clonidine on pulsatile luteinizing hor-mone secretion in a double blind study in men. J Clin

Endocrinol Metab 68:219 –222.

Klaiber EL, Broverman DM, Vogel W, Kobayashi Y, Moriarty D (1972): Effects of estrogen therapy on plasma MAO activity and EEG driving responses of depressed women. Am J

Psychiatry 128:1492–1498.

Klaiber EL, Broverman DM, Vogel W, Peterson LG, Snyder MB (1997): Relationships of serum estradiol levels, menopausal duration, and mood during hormonal replacement therapy.

Psychoneuroendocrinology 22:549 –58.

Li G, Regunathan S, Barrow CJ, Eshraghi J, Cooper R, Reis DJ (1994): Agmatine: An endogenous clonidine-displacing sub-stance in the brain. Science 263:966 –969.

McPherson CA (1985): Analysis of radioligand binding experi-ments. A collection of computer programs for the IBM PC.

J Pharmacol Methods 14:213–228.

Mishra N, Hamilton CA, Jones CR, Leslie C, Reid JL (1985): Alpha-adrenoceptor changes after oestrogen treatment in platelets and other tissues in female rabbits. Clin Sci 69:235– 238.

Orensanz LM, Guillamon A, Ambrosio E, Segovia S, Azuara MC (1982): Sex differences ina-adrenergic receptors in the rat brain. Neurosci Lett 30:275–278.

Parsons WB Jr, Morledge JH (1970): Antihypertensive effect of a new imidazoline compound (clonidine) and chlorthalidone, individually and in combination. Am J Cardiol 26:258 –261. Piletz J, Halaris A, Saran A, Marler M (1990): Elevated3

H-para-aminoclonidine binding to platelet purified plasma mem-branes from depressed patients. Neuropsychopharmacology 3:201–210.

Piletz JE, Andrew M, Zhu H, Feng YZ, Rains J, Halaris A (1998): Alpha2-adrenoceptors and I1-imidazoline binding

sites: Relationship with catecholamines in women of repro-ductive age. J Psychiatr Res 32:55– 64.

Piletz JE, Chikkala DN, Ernsberger P (1995): Comparison of the properties of agmatine and endogenous clonidine-displacing substance at imidazoline anda2adrenergic receptors. J

Phar-macol Exp Ther 272:581–587.

Piletz JE, Halaris A, Ernsberger PR (1994): Psychopharmacol-ogy of imidazoline anda2-adrenergic receptors: Implications

for depression. Crit Rev Neurobiol 9:29 – 66.

Piletz JE, Halaris A, Nelson J, Qu Y, Bari M (1996a): Platelet I1-imidazoline binding sites are elevated in depression but not

generalized anxiety disorder. J Psychiatr Res 30:147–168. Piletz JE, Halaris A, Saran A, Marler MR (1991): Desipramine

lowers 3

H-para-aminoclonidine binding in platelets of de-pressed patients. Arch Gen Psychiatry 48:813– 820. Piletz JE, Halaris AE, Chikkala D, Qu YS (1996b): Platelet I-1

imidazoline binding sites are decreased by two dissimilar

antidepressant agents in depressed patients. J Psychiatr Res 30:169 –184.

Piletz JE, Sletten K (1993): Nonadrenergic imidazoline binding sites on human platelets. J Pharmacol Exp Ther 267:1493– 1502.

Raddatz R, Lanier SM (1997): Relationship between imidazo-line/guanidinium receptive sites and monoamine oxidase A and B. Neurochem Int 30:109 –117.

Raddatz R, Parini A, Lanier SM (1997): Localization of the imidazoline binding domain on monoamine oxidase B. Mol

Pharmacol 52:549 –553.

Reis D, Regunathan S (1999): Agmatine: An endogenous ligand at imidazoline receptors is a novel neurotransmitter. Ann N Y

Acad Sci 881:65– 80.

Roberts JM, Goldfien RD, Tsuchiya AM, Goldfien A, Insel PA (1979): Estrogen treatment decreases a-adrenergic binding sites on rabbit pletelets. Endocrinology 104:722–728. Sherwin B (1988): Estrogen and/or androgen replacement

ther-apy and cognitive functioning in surgically menopausal women. Psychoneuroendocrinology 13:345–357.

Siever LJ, Uhde TW (1984): New studies and prespectives on the noradrenergic receptor system in depression: Effects of thea2

adrenergic agonist clonidine. Biol Psychiatry 19:131–156. Spitzer RL, Endicott J, Robbins E (1978): Research diagnostic

criteria: Rationale and reliability. Arch Gen Psychiatry 35: 773–782.

Stampfer MJ, Colditz GA, Willett WC, Manson JE, Rosner B, Speizer FE, Hennekens CH (1991): Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up from the nurses’ health study. N Engl J Med 325:756 –762. Stowell LI, McIntosh CJ, Cooke R, Ellis PM (1988):

Adreno-ceptor and imipramine reAdreno-ceptor binding during the menstrual cycle. Acta Psychiatr Scand 78:366 –368.

Sun M-K, Regunathan S, Reis DJ (1995): Cardiovascular re-sponses to agmatine, a clonidine-displacing substance, in anesthetized rat. Clin Exp Hypertens 17:115–128.

Sundaresan PR, Madan MK, Kelvie SL, Weintraub M (1985): Platelet a2 adrenoceptors and the menstrual cycle. Clin

Pharmacol Ther 37:337–342.

Tesson F, Limon-Boulez I, Urban P, Puype M, Vanderkerckhove J, Coupry I, et al (1995): Localization of I2-imidazoline

binding sites on monoamine oxidases. J Biol Chem 270: 9856 –9861.

Theodorou AE, Mistry H, Davies SL, Yamaguchi Y, Horton RW (1987): Plateleta2adrenoceptor binding and function during

the menstrual cycle. J Psychiatr Res 21:163–169.

U’Prichard DC, Snyder SH (1979): Distinct a-noradrenergic receptors differentiated by binding and physiological relation-ships. Life Sci 24:79 – 88.

Yaffe K, Grady D, Pressman A, Cummings S (1998): Serum estrogen levels, cognitive performance, and risk of cognitive decline in older community women [see comments]. J Am