Antidepressants

Reduce

Whole-Body

Norepinephrine

Turnover

While

Enhancing

6-Hydroxymelatonin Output

Robert

N.

Golden, MD;

Sanford

P.

Markey,

PhD;

Emile D.

Risby,

MD;

Matthew V.

Rudorfer,

MD;

Rex W.

Cowdry,

MD;

William Z.

Potter,

MD,

PhD

\s=b\

The effects of

antidepressant

treatment

on

noradrenergic

function

were

studied

in 27

patients

with

a

major

affective

disorder.

Twenty-four-hour

urinary

excretion

of

6-hydroxy-

melatonin and

"whole-body norepinephrine (NE)

turnover," ie,

24-hour

urinary

output

of NE and

its

major

metabolites

3-methoxy-4-hydroxyphenylglycol,

vanillylmandelic acid,

and

normetanephrine,

were

measured before and after

treatment

with the

tricyclic desipramine hydrochloride,

the aminoketone

bupropion hydrochloride,

the

nonselective

monoamine

oxi-

dase

(MAO)

inhibitor

tranylcypromine

sulfate,

and the

specific

MAO

type

A

inhibitor

clorgiline.

6-Hydroxymelatonin

excretion

increased

following antidepressant

treatment,

while at the

same

time

whole-body

NE turnover

was

reduced.

These find-

ings

support

the

hypothesis

that

antidepressant therapy

increases

noradrenergic "efficiency,"

in that functional

output,

as

measured

by

6-hydroxymelatonin,

is maintained while total

NE

production

is decreased.

(Arch

Gen

Psychiatry

1988;45:150-154)

For

more

than

two

decades, noradrenergic

systems

have

been

a

central

focus in theories

relating

to

the

patho¬

physiology

of

depression

and the mechanism of action of

antidepressants.

The

original

formulations1*3 of the

biogenic

amine

hypothesis

of

depression

were

advanced

at

a

time in

which

emerging technology permitted

the

measurement

of

a

major

norepinephrine

(NE)

metabolite,

3-methoxy-4-

hydroxyphenylglycol

(MHPG),

in

body

fluids.

Also,

the

first-generation antidepressant

medications shared

the

common

property

of

having

acute

biochemical effects

on

NE: monoamine

oxidase

(MAO)

inhibitors affect the

en¬

zyme

responsible

for the intraneuronal

degradation

of

NE;

secondary

amine

tricyclic antidepressants

block

reuptake

of NE

following

its

extraneuronal

release;

tertiary

amine

tricyclic antidepressants

are

metabolized

to

secondary

amines and

consequently

also affect

NE

reuptake.4·5

More

recent

observations have called into

question

the

central role of NE in

antidepressant

treatments.

Newer

agents

that do

not

have

acute

biochemical effects

on

NE

have

been

used

to treat

depression (eg, specific

serotonin

reuptake

inhibitors such

as

citalopram

and

zimelidine)/'

With the

recognition

that down

regulation

of

noradrenergic

receptors

following antidepressant

administration coin¬

cides

temporally

with the time

course

of clinical

response,

some

researchers have

suggested

that decreased noradre¬

nergic

functioning

may

be

an

essential factor

in

antide¬

pressant

effects.79

There is

a

growing appreciation

for the

dynamic

nature

of neurotransmitter

systems.

Thus,

it becomes difficult

to

interpret

the

physiologic meaning

of

a

static

measurement

of

a

single

component

of

the

system

in

any

given

compart¬

ment.

For

example,

if

low levels of MHPG

were

to

be found

in

the

cerebrospinal

fluid of

a

group

of

patients,

would this

represent

a

primary

event

(eg,

decreased

output

of NE

and

a

resultant decrease in

noradrenergic

function)

or a

secondary

compensatory

event

(eg,

response to

increased

receptor

sensitivity

and/or

density

in

a

hyperactive

sys¬

tem)?

Limitations

in

interpreting

monoamine metabolite

tissue concentrations have stimulated

a

search for

mean¬

ingful physiologic

indexes of

noradrenergic functioning.

Melatonin formation

may prove to

be

a

useful tool in this

regard.

As reviewed

by Lewy,10

several factors

suggest

that

pineal gland

secretion of melatonin could be utilized

as an

index of

noradrenergic

tone

in

man.

The

pineal gland

is

unique

in

that,

unlike other endocrine

organs,

it is

regulated primarily by

neural

(sympathetic)

innervation

and does

not

appear

to

be affected

by circulating

sub¬

stances.

Parasympathetic

innervation does

not

seem

to

be

Accepted

for

publication July

20,

1987.

From the

Laboratory

of Clinical Science

(Drs

Golden, Markey,

Risby,

Rudorfer,

Cowdry,

and

Potter)

and the

Office of the Clinical Director

(Dr

Cowdry),

National Institute of Mental

Health, Bethesda,

Md. Dr Golden is

now

with

the

University

of Mental

Health,

Bethesda,

Md. Dr Golden is

now

with

the

University

of North Carolina School of

Medicine,

Chapel

Hill.

Read

in

part

before the 137th annual

meeting

of the American

Psychiatric

Association,

Los

Angeles, May

8, 1984;

and

at

the 87th annual

meeting

of

the American

Society

for Clinical

Pharmacology

and

Therapeutics,

Wash-

ington,

DC,

March

20,

1986.

Reprint

requests

to

Department

of

Psychiatry, University

of North

Carolina

School of

Medicine,

Chapel Hill,

NC 27514

(Dr

Golden).

Downloaded From: http://archpsyc.jamanetwork.com/ by a Monash University Library User  on 04/14/2016

involved

in

melatonin

production.

Stimulation of

ß-adre¬

nergic

receptors

of

pinealocytes

sets

off

a

cascade of

events

leading

to

melatonin

production.11

a-Adrenergic

stimula¬

tion

potentiates

the

effects

of

ß-receptor

stimulation.1214

Pinealectomy

abolishes

plasma

levels of

melatonin,15

and

there is

no

substantial

evidence for

extrapineal production

of melatonin in

humans.16

Thus,

pineal gland

melatonin

formation should

provide

a

physiologic

marker for norad¬

renergic activity

and

potentially

could

serve as a means

for

examining

the

effects

of

antidepressant

treatment

on

noradrenergic

function.

Several

investigators

have measured

plasma

melatonin

levels in studies of

antidepressant

treatments

using

re¬

peated samples

over

all

or

part

of

a

24-hour

period.1720 By

calculating

the

area

under the

curve

for

repeated plasma

samples,

one can

compare melatonin

production

under

different conditions.

Alternatively, urinary

excretion of the

principal

metabolite of

melatonin, 6-hydroxymelatonin,

may

provide

a

convenient

means

for

assessing

melatonin

production.

Approximately

85% of melatonin is metabo¬

lized and

excreted

as

the

glucuronidated

and sulfated

conjugates

of

6-hydroxymelatonin. By measuring

total

6-hydroxymelatonin

output

in

a

24-hour urine

sample,

one

obtains

an

integrated

measure

of

melatonin formation.21

We have

applied

this tool

to

our

ongoing investigations

of the

mechanism of action of

antidepressant

treatments.

Previous work from

our

group has shown that diverse

therapies (electroconvulsive;

lithium

carbonate,

the

NE

reuptake

inhibitor

desipramine,

the serotonin

reuptake

inhibitor

zimelidine,

the

MAO

type

A

inhibitor

clorgiline,

and

the aminoketone

antidepressant bupropion)

all reduce

whole-body

NE

output

as

measured

by

the

summated

excretion of NE

plus

its

principal

metabolites

MHPG,

vanillylmandelic

acid

(VMA),

and

normetanephrine.2226

At

the

same

time

cardiovascular

functioning

is maintained

or

enhanced.27 This

can

be

interpreted

as

reflecting

increased

"efficiency"

in

NE

functioning

following antidepressant

therapy;

ie,

"less"

NE

is

doing

the

same or

more.27

To

test

this

hypothesis

further,

we

decided

to

measure

6-hydrox¬

ymelatonin

excretion and

whole-body

NE turnover

(   )

in

depressed patients

before and after

treatment.

If anti¬

depressants

in fact

increase

the

"efficiency"

of noradre¬

nergic

systems,

then

6-hydroxymelatonin

excretion should

be maintained in the face of

decreasing

total NE

produc¬

tion. We

report

herein

our

results with three

types

of

antidepressants:

monoamine oxidase

inhibitors

(tranylcy¬

promine

sulfate and

clorgiline),

a

tricyclic (desipramine

hydrochloride),

and

a

unicyclic

aminoketone

(bupropion

hydrochloride).

SUBJECTS

AND METHODS

Twenty-seven patients

who

fulfilled

Research

Diagnostic

Criteria28 for

major depression

were

admitted

to

a

clinical research

ward at

the National

Institutes of

Health, Bethesda,

Md. All

patients

were

placed

on

a

standard

low-monoamine,

restricted-

caffeine diet.29

Following

a

drug-free

washout

period

of

at

least

three

weeks,

baseline 24-hour

urine

samples

(7

am

to 7

am)

were

collected into bottles

containing

10 mL

of

a

3%

solution of

sodium

metabisulfite. The urine

samples

were

refrigerated

at

4°C

during

the

day

of collection.

After

measuring

the volume

of each

com¬

pleted

collection,

aliquote

were

frozen

at

-

20°C until the

time of

the assay.

Urine volumes

in

excess

of 900 mL

were

considered

acceptable

for

the

study.

The

lighting

conditions

on

the

ward did

not

vary

over

the

course

of

the

study,

and

patients

did

not

leave

the

hospital

on

the

days

in

which the 24-hour

urine

samples

were

collected.

Also, patients

did

not

engage in

excessive

or

strenuous

physical activity

during sample

collection

days. During sampling

periods,

if

patients

were

unable

to

sleep

at

night, they

were

encouraged

to

remain

resting

in

bed.

Under double-blind

conditions,

patients

were

assigned

to

receive

one

of the

following

treatments,

after

we

had obtained

informed

consent:

(1)

the

MAO inhibitor

tranylcypromine;

(2)

the

MAO

type

A inhibitor

clorgiline;

(3)

the

tricyclic antidepressant desipramine;

or

(4)

the

unicyclic

antidepressant bupropion.

The

seven

patients

treated with

bupropion

were

drawn from another

study,26

in

which

XNE,

but

not

6-hydroxymelatonin

output,

was

examined.

Dosage

was

titrated

as

tolerated under the

supervision

of

a

psychiatrist

who

was

not

"blind"

to

treatment. Table

1

lists the final dose

achieved

as

well

as

demographic

and

diagnostic

data for each

patient. Following

three

to

six weeks of

treatment,

24-hour urine

samples

were

again

obtained

using

identical

procedure

as

de¬

scribed above.

For

one

subject

(B) posttreatment

whole-body

NE

measures

were

not

determined,

although 6-hydroxymelatonin

concentrations

were

measured;

for

a

different

subject

(AA),

posttreatment

6-hydroxymelatonin

level

could not

be determined.

All

laboratory

assays

were

performed

on

coded urine

samples

by personnel

who

were

blind

to

the clinical

status

and

treatment

condition

represented by

each

sample. 6-HydroxymeIatonin

was

measured

using

a

negative

chemical ionization-mass

spectroscopic

method

as

previously

described.30

Briefly,

an

internal

standard

of

tetradeutero-6-hydroxymelatonin

sulfate

was

added

to

3 mL of

urine,

the

conjugates

were

hydrolyzed enzymatically,

and the free

melatonin

metabolites

were

extracted into dichloromethane. After

reaction with

i-butyldimethylchlorosilone

and

pentafluoropro-

pionic anhydride,

the stable

product

was

partially purified

on

a

silica

gel column,

and

the

ratio

of deuterated

to

endogenous

6-

hydroxymelatonin

derivatives

was

determined

using

gas

chromat-

ographic-mass spectrometric analysis.

The

urine

samples

were

assayed

for NE

and its

major

metabolites

(MHPG,

VMA,

and

normetanephrine)

using

mass

fragmentography.31

The

sum

of

NE,

normetanephrine,

MHPG,

and

VMA

output

was

used

to

indicate

   .23

When

two

or

three consecutive

daily

urine

samples

could

be obtained from

a

patient during

the baseline

and/or

postantide-

pressant

treatment

conditions,

mean

values for

6-hydroxymela¬

tonin and

   

were

taken.

Baseline

and

posttreatment

samples

for each

patient

were

assayed

on

the

same run.

Pretreatment and

posttreatment

measures

were

compared

us¬

ing

t

tests

for related

samples.

To determine if the effects of

the

two

MAO

inhibitors

were

different,

a

Mann-Whitney

U

test

was

applied.

Since

clorgiline

and

tranylcypromine

did

not

have

differ¬

ent

effects

on

either

   

(P>.50)

or

6-hydroxymelatonin

(P

=

.50),

the

data from

patients receiving

MAO inhibitor

treat¬

ment

were

pooled

for

subsequent analyses.

RESULTS

Whole-body

NE

turnover

was

significantly

reduced

following

antidepressant

treatment,

falling

from

a mean

(

±

SEM)

pretreat¬

ment

level of

38.3±2.1

(6.5±0.4)

to

23.8±1.8

µ    /d

(4.0±

0.3

mg/24 h) (P

=

.0001;

Table

1).

(A

more

detailed table

showing

the effects of

treatment

on

urinary

NE and

each

of its

metabolites

is available

on

request.)

At the

same

time,

6-hydroxymelatonin

excretion increased from

a

mean

baseline of

7.2 ±1.2

µg/24

h to

a

mean

posttreatment

level of

9.2

±

1.4

p.g/24

h

=

.04;

Table

1).

The reduction

in

   

was

greatest

in

patients receiving

MAO

inhibitor

treatment,

all

of whom

demonstrated this effect

(P

=

.0003).

6-Hydroxymelatonin

excretion increased in

eight

of

ten

patients

so

treated,

a

trend that did

not

quite

reach

statistical

significance

(P

=

.08).

Desipramine

also

significantly

reduced

   

(P

=

.0009),

with

a

trend toward

increasing

6-hydroxymelatonin

excretion

(P

=

.14).

Bupropion

reduced

   

(P<.05);

its

effect

on

6-hydroxymelatonin

was

not

significant

(P

=

.80;

Table

2).

COMMENT

These

findings

confirm

previous

reports

that

antidepres¬

sant treatments

lead

to

a

reduction in

   

in

depressed

patients.22*26

At the

same

time,

we

found that

6-hydroxy¬

melatonin excretion increased.

Thus,

our

study

fails to

provide

evidence for

a

reduction in

noradrenergic

function

following long-term antidepressant

treatments.

On the

contrary,

our

findings

are

consistent with

an

interpretation

that

antidepressant

therapy

leads

to

an

increase

in

norad¬

renergic

"efficiency,"

in

that functional

output

as

previously

Downloaded From: http://archpsyc.jamanetwork.com/ by a Monash University Library User  on 04/14/2016

Table

1.—Demographic, Diagnostic,

and

Treatment Data*

Patient/Age, y/Sex

RDC

Diagnosis

Final

Dose,

mg/d

6-Hydroxy

melatonin,

Kg

24

h

   ,

µ    

d

(mg/24

h)

Baseline

Posttreatment

Baseline

Posttreatment

A/51/F

UP

50

Tranylcypromine

Sulfate

1.2

6.1

42.6

(7.2

22.9

(3.9

B/36/M

UP

40

4.8

7.1

C/34/F

BP-II

40

2.0

2.5

42.9

7.3

11.8

2.0

D/70/F

UP

Clorgiline

1.5

5.2

23.9

4.0

16.0

2.7

E/56/M

BP-I

10

26.5

31.2

58.9

10

0)

29.6

5.0

F/30/M

BP-II

15

8.9

14.1

39.6

6.7

20.1

3.4

G/42/M

BP-I

15

8.2

6.7

67.0

11

3)

23.3

3.9

H/37/F

UP

10

14.9

9.8

33.0

5.6

13.2

2.2

I/32/M

BP-I

2.5

6.1

10.5

28.4

4.8

18.9

3.2

J/44/F

BP-I

7.3

9.3

39.7

6.7

12.8

2.2

K/29/F

UP

150

Desipramine

Hydrochloride

7.3

0.8

41.8

7.1

16.7

2.8

L/48/F

BP-II

250

0.6

6.4

49.2

8.3

31.3

5.3

M/35/F

UP

200

6.4

1.0

42.0

7.1

33.3

5.6

N/51/F

BP-I

250

2.5

2.7

34.4

5.8

29.7

5.0

O/40/F

UP

150

3.3

4.3

29.1

4.9

18.8

3.2

P/28/F

BP-I

225

11.3

15.9

28.6

4.8

16.4

2.8

Q/48/F

UP

200

5.9

14.0

28.2

4.8

9.2

1.6

R/32/M

UP

400

14.8

19.8

39.0

6.6

38.1

6.4

S/32/F

BP-II

200

3.2

23.1

3.9

18.8 3.2

T/69/F

UP

100

11.7

19.3

41.3

7.0

19.8

3.4

U/41/F

UP

300

Bupropion Hydrochloride

3.1

5.5

18.6

3.1

22.7

3.8

V/31/F

UP

400

18.1

12.2

31.4

5.3

23.9

4.0

W/51/F

BP-I

500

4.5

5.6

51.1

8.6

41.1

7.0

X/26/M

BP-I

425

6.2

17.2

39.8

6.7

40.6

6.9

Y/53/F

BP-I

375

6.9

1.3

43.3

7.3

28.6

4.8

Z/54/M

BP-II

400

1.0

2.0

39.3

6.6

33.5

5.7

AA/26/F

BP-I

400

40.7

(6.9

27.0

(4.6

41.8±2.3t

7.2±1.2t

3.2±1.4t

38.3±2.1

(6.5±0.4)t

23.8±1.8

(4.0±0.3)t

*RDC indicates Research

Diagnostic

Criteria;    ,

whole-body

norepinephrine

turnover; UP,

unipolar;

BP,

bipolar (I

and

II).

tMean±SEM.

Table

2.—Drug

Effects

on

6-Hydroxymelatonin

and

   *

6-Hydroxy

melatonin,

g

24

h

   , µ    /d

(mg/24 h)

treatment

Post-

Post-

Group

Baseline

treatment

Baseline

treatment

MAO inhibitor

8.1 ±2.4

10.3±2.5

41.8±4.6

18.7±2.0

(7.1 ±0.8)

(3.2±0.3)

Desipramine

6.7±1.5

9.2±2.3

35.7±2.6

23.2±2.9

hydrochloride

(6.0 ±0.4)

(3.9

±0.5)

Bupropion

6.6±2.5

7.3±2.5

37.7±3.9

31.1 ±2.9

hydrochloride

(6.4 ±0.7)

(5.3

±0.5)

*   

indicates

whole-body norepinephrine

turnover;

MAO,

monoamine

oxidase.

Values

are mean ±

SEM.

measured

by

cardiovascular

values27

and in the

present

instance

by 6-hydroxymelatonin

excretion is

not

only

maintained but is

possibly

enhanced,

while

at

the

same

time

total

NE

production

is

decreased.

Could

each of the observed

antidepressant

effects

simply

reflect

the

result of

drug-induced changes

in the metabo-

lism of NE? For

example, by decreasing presynaptic

reuptake

of NE and

consequent

intraneuronal

catabolism,

NE metabolite

formation could

be

altered.

Also,

the

reuptake

process

serves

physiologically

to

protect

the

pineal

gland

from

nonsynaptic adrenergic

stimulation;

reuptake

blockade

can

increase the response of

pineal

N-acetyltransferase activity

to

stimulation

by

stress.32

However,

drug-induced changes

in metabolism

are

unlikely

to

explain

our

findings completely

since

we

included differ¬

ent

types

of

pharmacologie

agents

with

markedly

different

effects

on

catecholamine metabolism.

Preclinical studies

examining

the effects of

antidepres¬

sant

administration

on

pineal

function

in

animals have

yielded apparently conflicting

results.

King

et

al33

reported

that

MAO inhibition led

to

an

increase in melatonin

content

of

rat

pineal

gland. Oxenkrug

and coworkers34 also found

an

increase in

rat

pineal

melatonin

content

following

administration of

an

MAO

type

A

inhibitor

but

not

with

an

MAO

type

 

inhibitor;

the former effect

was

greatly

diminished

by superior

cervical

ganglionectomy35 Heydorn

et

al,36

on

the other

hand,

looking

at

stimulated melatonin

output

instead of

pineal

basal

content,

found that

long-

term,

though

not

short-term,

desipramine

or

nialamide

Downloaded From: http://archpsyc.jamanetwork.com/ by a Monash University Library User  on 04/14/2016

exposure

reduced both the

isoproterenol-

and darkness-

induced elevation of melatonin in

rat

pineal gland

and

serum.

The

same

group,

however,

has

recently reported

that

long-term desipramine

treatment

does

not

decrease

nocturnal

plasma

melatonin concentrations in

depressed

patients.18

Our

findings

of increased

urinary

6-hydroxymelatonin

excretion

following antidepressant therapy

are

compatible

with this and other

recent

reports

of the effects of antide¬

pressants

on

plasma

melatonin

in

man.

Thompson

et

al19

studied

six

normal

subjects

and

six

depressed patients

following long-term desipramine

treatment.

The

depressed

subjects

showed

a

significant

rise in nocturnal melatonin

secretion after three weeks of

desipramine

treatment.

Interestingly,

the normal

subjects

did

not

demonstrate

significant

differences in

pretreatment

and

posttreatment

nocturnal

plasma

melatonin concentrations. Cowen

et

al20

studied

ten

healthy subjects

who

were

given desipramine

and found that short-term administration led

to

a

signifi¬

cant

increase in

mean

midnight plasma

melatonin

concen¬

trations. This increase

peaked

on

the fifth

day

of

desipra¬

mine

administration

and

then

returned

toward

pretreatment

levels with

no

significant

difference between

baseline and

19-day

treatment

values.

Murphy

et

al37

examined the effects of three different MAO inhibitors

on

morning plasma

melatonin levels in

depressed patients.

Both

clorgiline

and

tranylcypromine,

the MAO

type

A

inhibitor and the nonselective MAO inhibitor used in

our

study,

increased

morning

plasma

melatonin concentrations

in

depressed patients following long-term

(three-week)

treatment;

the

MAO

type

 

inhibitor

deprenyl (selegiline),

in

contrast,

did

not

increase

plasma

melatonin

levels,

suggesting

that

inhibition

of MAO

type

A

by

either

a

selective

or a

nonselective MAO inhibitor

can

increase

pineal

melatonin

output.

Mendlewicz

et

al17

reported

that

in four

depressed patients,

mean

plasma

melatonin values

and

day-night

differences

were

similar before and after

four weeks'

treatment

with

amitriptyline.

In the

only

other

published study examining

the effects of

antidepressant

treatment

on

urinary

excretion of

6-hydroxymelatonin,

Sack and

Lewy38 reported

that in four

depressed patients,

desipramine

treatment

led

to

an

increase in

6-hydroxy¬

melatonin

output,

which

was

sustained

over

three weeks

of

treatment.

Despite

the overall

general

agreement,

it is

not

clear

from the

studies

of either

plasma

melatonin

or

urinary

6-hydroxymelatonin

that

a

stable,

reproducible drug-in¬

duced increment

can

be shown after

treatment

with MAO

inhibitors

(including

type

A

inhibitors)

and

tricyclic

anti¬

depressants.

Our

sample

size for each

treatment group

is

relatively small,

and the increase in

6-hydroxymelatonin

excretion

following antidepressant

treatment

reaches

sta¬

tistical

significance only

when the three

treatment groups,

including bupropion,

are

pooled.

At the time that this

study

was

designed

we

believed that

including

bupropion

as one

of the

treatments

would enable

us

to

look

at

the

effects

on

noradrenergic functioning

of

a

drug

that did

not

have direct biochemical effects

on

this

system.

Though

bupropion

is characterized

preclinically

as

a

dopamine

uptake

inhibitor without

appreciable

effects

on

NE

reup¬

take

or on

MAO,6

one

of the

drug's principal

metabolites in

man,

hydroxybupropion,

may

in fact be

an

NE

reuptake

inhibitor.39 It is

possible,

therefore,

that

bupropion,

through

its active

metabolite,

is

exerting desipraminelike

effects

on

NE

uptake.40

Even

though

the

antidepressant-induced

changes

in

melatonin

measures are

not

always

robust and

in

the

same

direction,

our

studies do reveal

remarkably

consistent

treatment-associated reductions

in

whole-body

NE

output.

Only

two

of 27

patients (patients

U

and

X)

failed

to

show

the reduction in

   

(Table 1).

Since the

same

urine

samples

were

used

to

measure

output

of both total NE and

6-hydroxymelatonin,

it is of interest

to

assess

the relation¬

ship

between the factors. In these

patients,

however,

all of

whom

were

depressed during

the baseline

measure,

no

correlation could be identified

using

standard

parametric

and

nonparametric

approaches.

Thus,

although

melatonin

production

may

be

extensively

controlled

by

the level of

stimulation of

ß-receptors

on

the

pineal gland,

it does

not

appear

to

relate

appreciably

to

total NE

output

in

de¬

pressed patients

in any

direct

manner.

These

findings

suggest

caution in

using

melatonin

output

as a

functional NE

measure.

It

might

be

argued

on

the

basis of the model discussed in the introduction that

the

variation in melatonin

or

6-hydroxymelatonin

is

somehow

more

directly

related

to

brain

noradrenergic

function

than

is

urinary

NE

output. However,

pineal

melatonin

produc¬

tion is

not

a measure

of

central

noradrenergic

function.

The

pineal gland

lies outside of the blood-brain

barrier41;

it

can

be affected

by circulating

catecholamines

when

reuptake

is blocked.32 Its innervation is via the

superior

cervical

ganglion,

and its

activity

does

not

necessarily

parallel

that of other

noradrenergic

systems

in

the

brain.

Also

arguing against

melatonin

production being

more

directly

related

to

central

noradrenergic

function than is

urinary

NE

output

is

an

extensive animal literature

show¬

ing

that

antidepressants produce

consistent reductions in

central

nervous

system

ß-adrenergic

receptors8*9;

the

con¬

comitant

finding

in

man

is the reduction of total

NE

output.42

We

believe

that this reduction

must

be of central

origin. Tricyclics

can

increase

plasma

NE

levels,43*44

pre¬

sumably through

reduction in clearance

(which

has

been

demonstrated in

humans45),

despite

a

reduction

in

total

output.22

Monoamine oxidase

inhibitors,

in

humans,

reduce

plasma

NE

levels,46*47

but in the absence of central influ¬

ences

in

the

rat

(ie,

following "pithing"), they

too

can

actually

increase

plasma

NE

levels.48 These

findings

are

compatible

with

the observations that locus ceruleus

firing

in

the central

nervous

system

of rodents is

markedly

reduced after both

tricyclic

and

MAO inhibitor administra¬

tion.49·50

Plasma melatonin and

urinary 6-hydroxymelatonin

mea¬

sures

are,

however,

useful

in

demonstrating

that

no

matter

how

great

the

antidepressant-induced

down

regulation

of

ß-receptors,

the

output

is

not

reduced. If

anything,

chronic

antidepressant

use

enhances melatonin

output.

This is

compatible

with Stone's61

suggestion

that

drug

effects such

as

down

regulation

of

ß-receptors

may

be

partially

com¬

pensatory

to

and of less functional

significance

than the

opposing

forces of

uptake

inhibition

and MAO

inhibition.

These

findings

support

the

theory

that

antidepressant

treatment

leads

to

increased

"efficiency"

in

noradrenergic

systems

with

an

enhancement

or

maintenance of

function

and

a

decrease in total NE

production.

References

1.

Prange

AJ Jr: The

pharmacology

and

biochemistry

of

depression.

Dis

Nerv

Syst

1964;25:217-221.

2.

Schildkraut JJ: The catecholamine

hypothesis

of affective

disorders:

A

review of

supporting

evidence. Am J

Psychiatry

1965;122:509-522.

3.

Bunney

WE

Jr,

Davis JM:

Norepinephrine

in

depressive

reactions: A

review. Arch

Gen

Psychiatry

1965;13:483-494.

4. Carlsson

A,

Corrodi

H,

Fuxe

K,

Hokfelt T: Effects of

some

antide-

pressant

drugs

on

the

depletion

of intraneuronal brain catecholamine

stores

caused

by

4-alpha-dimethyl-meta-tyramine.

Eur

J Pharmacol

1969;5:367\x=req-\

373.

5.

Glowinski

J,

Axelrod J: Inhibition of

uptake

of tritiated noradrenaline

in

the intact

rat

brain

by

imipramine

and

structurally

related

compounds.

Downloaded From: http://archpsyc.jamanetwork.com/ by a Monash University Library User  on 04/14/2016

Nature

1964;204:1318-1319.

6.

Rudorfer

MV,

Golden

RN,

Potter WZ: Second

generation antidepres-

sants.

Psychiatr

Clin

North Am

1984;7:519-534.

7.

Vetulani

J,

Stamarz

R,

Dingell

JW,

Sulser F: A

possible

common

mechanism

of action of

antidepressant

treatments:

Reduction in the sensi-

tivity

of the

noradrenergic

cycle

AMP

generating

system

in the

rat

limbic

forebrain.

Naunyn

Schmiedebergs

Arch Pharmacol

1976;239:109-114.

8.

Sulser

F,

Vetulani

J,

Mobley

P:

Mode

of action of

antidepressant

drugs.

Biochem Pharmacol

1978;27:257-261.

9.

Charney

DS,

Menkes

DB,

Heninger

GR:

Receptor

sensitivity

and

the

mechanism of action of

antidepressant

treatment.

Arch

Gen

Psychiatry

1981;38:1160-1180.

10.

Lewy

AJ: Human melatonin secretion: I. A

marker

for

adrenergic

function,

in Post

RM,

Ballenger

JC

(eds):

Neurobiology of

Mood Disorders.

Baltimore,

Williams &

Wilkins, 1984,

pp 207-214.

11.

Axelrod J: The

pineal gland:

A

neurochemical transducer. Science

1974;184:1341-1348.

12. Klein

DC,

Sugden D,

Weller JL:

Post-synaptic alpha-adrenergic

receptors

potentiate

the

beta-adrenergic

stimulation of

pineal

serotonin

N\x=req-\

acetyltransferase.

Proc Natl Acad

Sci USA

1983;80:599-603.

13.

Sugden

D,

Weller

JL,

Klein

DC,

Kirk

KL:

Creveling

R:

Alpha-

adrenergic

potentiation

of

beta-adrenergic

stimulation of

rat

pineal

N-

acetyltransferase.

Biochem Pharmacol 1984;33:3947-3950.

14.

Sugden

D,

Namboodiri

MAA,

Klein

DC,

Pierce

JE,

Grady

R

Jr,

Mefford IN: Ovine

pineal alpha1-adrenoceptors:

Characterization and

evi-

dence for

a

functional role in the

regulation

of

serum

melatonin. Endocri-

nology

1985;116:1960-1967.

15.

Lewy

AJ,

Tetsuo

M,

Markey

SP,

Goodwin

FK,

Kopin

IJ:

Pinealectomy

abolishes

plasma

melatonin

in the rat.

J Clin Endocrinol Metab

1980;50:204\x=req-\

205.

16.

Wetterberg

L:

Melatonin in humans:

Physiological

and clinical

stud-

ies.

J Neural

Transm

1978;13(suppl):289-310.

17. Mendlewicz

J,

Branchey

L,

Weinberg V,

Branchey

M,

Linkowski

P,

Weitzman ED: The 24-hour

pattern

of

plasma

melatonin in

depressed

patients

before and after

treatment.

Commun

Psychopharmacol

1980;4:49\x=req-\

55.

18.

Gottlieb

J,

Amsterdam

J,

Caroff

S,

Winokur

A,

Frazer

A:

Serum

melatonin

in

depressed

patients

before and after

treatment

with

desipra-

mine,

abstracted. Soc Neurosci Abstr

1984;10:263.

19.

Thompson

C,

Mezey

G,

Corn

T,

Graney

C,

English

J,

Arendt

J,

Checkley

SA: The effect of

desipramine

upon

melatonin and cortisol

secretion in

depressed

and normal

subjects.

Br

J

Psychiatry 1985;147:389\x=req-\

393.

20.

Cowen

PJ,

Green

AR,

Graham-Smith

DG,

Braddock LE: Plasma

melatonin

during desmethylimipramine

treatment:

Evidence for

changes

in

noradrenergic

transmission.

Br

J Clin Pharmacol

1985;19:799-805.

21.

Markey

SP,

Higa S,

Shih

M,

Danforth

DN,

Tamarkin

L:

The

correlation between human

plasma

melatonin levels and

urinary 6-hydroxy-

melatonin excretion. Clin Chim Acta

1985;150:221-225.

22.

Linnoila

M,

Karoum

F,

Calil

H,

Kopin

IJ,

Potter WZ: Alteration of

norepinephrine

metabolism with

desipramine

and zimelidine in

depressed

patients.

Arch Gen

Psychiatry

1982;39:1025-1028.

23.

Linnoila

M,

Karoum

F,

Potter WZ: Effect of low-dose

clorgyline

on

24-hour

urinary

monoamine excretion in

patients

with

rapidly

cycling

bipolar

affective disorder.

Arch Gen

Psychiatry

1982;39:513-516.

24.

Linnoila

M,

Karoum

F,

Rosenthal

N,

Potter WZ: Electroconvulsive

treatment

and lithium carbonate: Their effects

on

norepinephrine

metabo-

lism in

patients

with

primary

major

depressions.

Arch

Gen

Psychiatry

1983;40:677-680.

25.

Rudorfer

MV,

Scheinin

M,

Karoum

F,

Ross

RJ,

Potter

WZ,

Linnoila

M: Reduction of

norepinephrine

by serotonergic

drug

in

man.

Biol

Psychi-

atry

1984;19:179-193.

26.

Golden

RN,

Rudorfer

MV,

Sherer

MA,

Linnoila

M,

Potter

WZ:

Bupropion

in

depression:

I.

Biochemical effects and clinical response. Arch

Gen

Psychiatry,

1988;45:139-143.

27.

Potter

WZ,

Rudorfer

MV,

Linnoila M: New clinical studies

support

a

role of

norepinephrine

in

antidepressant

action,

in Barchas

JB,

Bunney

WE

Jr

(eds):

Perspectives

in

Psychopharmacology.

New

York,

Alan R Liss

Inc,

in press.

28.

Spitzer

RL,

Endicott

J,

Robins E: Research

Diagnostic

Criteria

for

a

Selected

Group of

Functional

Disorders,

ed 2. New

York,

New York

State

Psychiatric

Institute,

1981.

29. Muscettola

G,

Wehr

T,

Goodwin F: Effect of diet

on

urinary

MHPG

excretion in

depressed patients

and in normal controls.

Am

J

Psychiatry

1977;134:914-916.

30.

Tetsuo

M,

Markey

SP,

Colburn

RW,

Kopin

IJ:

Quantitative analysis

of

6-hydroxymelatonin

in human urine

by

gas

chromatography-negative

ionization

mass

spectrometry.

Anal Biochem

1981;110:208-215.

31. Karoum

F,

Neff NH:

Quantitative

gas

-chromatograpy\p=m-\mass

spec-

trometry (GC-MS)

of

biogenic

amines:

Theory

and

practice,

in

Spector S,

Beck N

(eds):

Modern Methods

in

Pharmacology.

New

York,

Alan R Liss

Inc,

1982,

pp

39-54.

32. Parfitt

AG,

Klein DC:

Sympathetic

nerve

endings

in the

pineal gland

protect

against

acute

stress-induced increase

in

N-acetyltransferase

(EC

2.3.1.5)

activity.

Endocrinology

1976;99:840-851.

33.

King

T,

Richardson

D,

Reiter

R:

Regulation

of

rat

pineal

melatonin

synthesis:

Effect of monoamine oxidase inhibition. Mol Cell Endocrinol

1982;25:327-338.

34.

Oxenkrug

GF,

McCauley R, McIntyre

I,

Filipowicz

C: Selective

inhibition of MAO-A but

not

MAO

B

activity

increases

rat

pineal

melatonin.

J Neural

Transm

1985;61:265-270.

35.

McIntyre

IM,

McCauley

R,

Murphy

S, Goldman,H,

Oxenkrug

GF:

Effect of cervical

ganglionectomy

on

melatonin stimulation

by specific

MAO-A inhibition. Biochem Pharmacol

1985;34:3393-3394.

36.

Heydorn

WE,

Brunswick

DJ,

Frazer A: Effect

of treatment

of

rats

with

antidepressants

on

melatonin concentrations in the

pineal gland

and

serum.

J Pharmacol

Exp

Ther

1982;222:534-543.

37.

Murphy

DL,

Tamarkin

L,

Sunderland

T,

Garrick

N,

Cohen RM:

Human

plasma

melatonin is elevated

during

treatment

with the

monoamine

oxidase inhibitors

clorgyline

and

tranylcypromine

but

not

deprenyl. Psy-

chiatry

Res

1986;17:119-128.

38. Sack

RL, Lewy

AJ:

Desmethylimipramine

treatment

increases

melatonin

production

in humans. Biol

Psychiatry

1986;21:406-410.

39.

Perumal

AS,

Smith

TM,

Suckow

RF,

Cooper

TB:

Effect of

plasma

from

patients

containing bupropion

and

its

metabolites

on

the

uptake

of

norepinephrine. Neuropharmacol

1986;25:199-202.

40.

Golden

RN,

DeVane

L,

Laizure

SC,

Rudorfer

MV,

Sherer

M,

Potter

WZ:

Bupropion

in

depression:

II.

The role of metabolites

in

clinical

outcome.

Arch Gen

Psychiatry, 1988;45:145-149.

41.

Reiter RJ: The mammalian

pineal gland:

Structure and function. Am

J

Anat

1981;162:287-313.

42.

Golden

RN,

Potter

WZ:

Neurochemical and neuroendocrine

dysregu-

lation in affective disorders.

Psychiatr

Clin North

Am

1986;9:313

327.

43. Ross

RJ,

Zavadil AP

III,

Calil

HM,

Linnoila

M,

Kitanaka

I,

Blomberg

P,

Kopin

IJ,

Potter WZ: Effects of

desmethylimipramine

on

plasma norepi-

nephrine, pulse,

and blood pressure. Clin Pharmacol Ther

1983;33:429\x=req-\

437.

44.

Veith

RC,

Raskind

MA,

Barnes

RF,

Gumbrect

G,

Ritchie

JL,

Halter

JB:

Tricyclic

antidepressants

and

supine, standing,

and exercise

plasma

norepinephrine

levels. Clin Pharmacol Ther

1983;33:763-769.

45.

Esler

M,

Turbott

J,

Schwarz

R,

Leonard

P,

Bobik

A,

Skews

H,

Jackman G: The

peripheral

kinetics of

norepinephrine

in

depressive

illness.

Arch Gen

Psychiatry

1982;39:292-300.

46. Ross

RJ,

Scheinin

M,

Lesieur

P,

Rudorfer

MV, Hauger RL,

Siever

LJ,

Linnoila

M,

Potter WZ: The effect of

clorgyline

on

noradrenergic

function.

Psychopharmacology

1985;85:227-230.

47.

Murphy

DL,

Lipper

S,

Campbell

IC,

Major

LF,

Slater

SL,

Bucks-

baum MS:

Comparative

studies of MAO-A and MAO-B inhibitors

in man,

in

Singer

TP,

Von Korff

RW,

Murphy

DL

(eds):

Monoamine Oxidase: Struc-

ture,

Function,

and Altered Function.

Orlando, Fla,

Academic Press

Inc,

1979,

pp

457-475.

48.

Finberg

JPM,

Kopin

IJ: Chronic

clorgyline

treatment enhances

release of

norepinephrine following sympathetic

stimulation in

the

rat.

Naunyn Schmiedebergs

Arch Pharmacol

1986;332:236-242.

49. Blier

P,

DeMontigny

C,

Azzaro J: Modification of

serotonergic

and

noradrenergic

neurotransmission

by repeated

administration of monoamine

oxidase inhibitors:

Electrophysiological

studies in the

rat

central

nervous

system.

J Pharmacol

Exp

Ther

1986;237:987-994.

50.

Nyback

HV,

Walters

JV,

Aghajanian

GK,

Roth RH:

Tricyclic

antide-

pressants:

Effects

on

firing

rate

of brain

noradrenergic

neurons.

Eur J

Pharmacol

1975;32:302-312.

51.

Stone EA: Problems with

current

catecholamine

hypotheses

of

antidepressant

agents:

Speculations leading

to

a

new

hypothesis.

Behav

Brain

Sci

1983;6:535-577.

Downloaded From: http://archpsyc.jamanetwork.com/ by a Monash University Library User  on 04/14/2016