Reprinted from

28 June 1974, Volume 184, pp. I341 - I348

SCIENCE

      The Pineal Gland:
A Neurochemical Transducer

  Chemical signals from nerves regulate synthesis of
melatonin and convey information about internal clocks.

Julius Axelrod

The pineal gland has become the
subject of considerable investigation
during the past decade because it pro-
vides a productive experimental model
for studying circadian rhythms and reg-
ulation of and organs by nerves. In the
mammal, the pineal gland rests between
the two cerebral hemispheres and
weighs about 100 milligrams in man
and 1 mg in the rat (I ). The pineal
gland originates in the brain of the
developing mammalian embryo, but it
loses direct nerve connection with the
brain soon aftor Mrth. The pineal paren-
chymal cells .are innervated by sympa-
thetic nervom (noradrenaline-containing)
wboso cell bodies lie in the superior
cervical ganglia (2). Amphibian pineals
llW0 photoroceptive cells that can gen-
orate nerve impulses in direct response
to environmental light (3). Photorecep-
tor elements, however, are not found in
the mammalian pineal cells.

The beginning of the modern era in
pineal research stemmed from the iso-
lation and identification of the indole-
amine melatonin (S-methoxy-N-acetyl-

Ihs author II chief of the pharmacology section,
kbDrrsorY of Cliaic~l Science, National Institute
Of bbttd &W&b, &them-la, Maryland 20014.

tryptamine) from bovine pineals by
Lerner et al. (4). It then became possi-
ble to examine its localization, physio-
logic properties, formation, and metab-
olism: Melatonin is the most potent
agent for causing contractions of me-
lanophores in frog and fish skin. When
treated with melatonin at concentrations
of IO-l3 gram per milliliter, the s&ii
of many fish and amphibians rapidly
blanches (5). The amphibian pineal
contains melatonin and the enzymes
that make it (6). These results indicate
that melatonin causes changes in skin
pigmentation in fish and amphibians
when it is released from pineal organs.
In the mammal, melatonin is synthe-
sized mainly in the pineal (I), and it
exerts inhibitory effects on gonads. When
injected into birds, it causes a decrease
in &weight of the ovaries, testes, and
oviduct (I). It delays vaginal opening
and reduces ovary weight in young
rats (7). When melatonin is implanted
in the median eminence, the elevation
in the content of leutinizing hormone
(LH) in the pituitary following castra-
tion is blocked, and plasma LH con-
centration is lowered (8). Blinding of
male hamsters causes a fall in the
weight of testes, but when .pineals are

removed or when nerves to the pineal
are cut the reduction in testicular
weight is preveqted. #During proestrus
in rats, melatonin inhibits ovulation by
preventing the release of LH (9). The
early morning elevation in plasma pro-
lactin in male rats is mediated by in-
creased release of a pineal hormone
(IO). In the sparrow, the pineal serves
as a time-measuring system (II ). The
physiological aspects of the pineal have
been reviewed recently (12).

Melatonin is synthesized almost exclu-
sively within the pineal cell as follows
(Fig. 1): tryptophan + S-hydroxytrypto-
phan + serotonin + N acetylserotonin +
melatonin. Tryptophan is hydroxyl-
ated to S-hydroxytrytophan by trypto-
phan hydroxylase (13). The _ latter
amino acid is then decarboxylated
by Z-aromatic amino acid decarboxylase
to form the biogenic amine serotonin.
Serotonin then undergoes a complex
fate. One portion is deaminated to 5-
hydroxyindoleacetic acid by mono-
amine oxidase, and another portion
leaves the pineal cell and is taken up
by (he sympathetic nerve terminal and
stored together with the neurotrans-
mitter noradrenaline (I) (Fig. 1). A
third portion is acetylated to N-acetyl-
serotonin ,by the enzyme serotonin N-
acetyltransferase (14). This is a critical
regulatory step, as will be shown later.
N-Acetylserotonin is then 0-methylated
by hydroxyindole 0-methyltransferase
to form melatonin, S-adenosylmethio-
nine serving as the methyl donor (IS).

Hydroxyindole 0-methyltransferase is
highly. localized in the pined glands of
mammals and birds. Small amounts of
the enzyme are also present in the
retina of the rat. In other classes (rep-
tiles, amphibia, and fish), hydroxy-
indole  0-methyltransferase is also
found in *the eye and brain as well as
the pineal region ,(.16). Although in-
direct, the evidence that the frog pineal
blanches skin by secreting melatonin is
compelling.

Copyright@ 1974 b1/ the American Asaociution for the Advancem.e& of Sc&m.ce

Reprinted by the

U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
        National Institutes of Health


NWVO

Lit and the Melatonin-Forming
Enzyme

Rats exposed to constant illumination
remain in persistent estrus. This condi-
tion could be reduced or prevented
when rats were injected with extracts
of bovine pineal glands (17). As a re-
sult of these findings Wutiman et al.
(17) concluded that the pineal gland
releases a substance that inhibits the
gonads and that the formation and
releasd of `this substance is reduced
when animals `are kept in constant illu-
mination. Fiske et al. (18) also found
that pineal glands of rats exposed to
continuous light weighed less. Then my
colleagues and I (7) found that mela-
tonin reduces the incidence of estrus in
rats exposed to continuous light. It be-
came apparent that environmental light-
ing might effect the melactonin-forming
enzyme, hydroxyindole O-methyltrans-
ferase. Rats .kept in continuous light
for about 7 days showed a marked re-
duction in enzyme activity as compared
to .those kept in constant darkness (19).
Thus, constant light decreased the ac-
tivity of hydroxyindole O-methyltrans-
ferase, which in turn reduced the
production  of the gonad-inhibiting
compound, melatonin. This reduction
of melatonin synthesis in constant light
would result in persistent es&us.

The question then arose as to how
messages abouxt environmental lighting
could reach the pineal, which lies deep
between the two cerebral hemispheres.
The most likely possibility was a neural
pathway. The mammalian pineal is
heavily  innervated by sympathetic
nerve terminals, whick are highly
branched and contain swellings or vari-
cosities that are in close juxtaposition
with pineal parenchymal cell? (Fig. 1).
These varicosities contain numerous
granulated vesicles that are the site of
storage of the neurotransmitter nor-
adrenaline (20). The neNe terminals
that innervate the pineal can readily
be destroyed by bilateral removal of
the superior cervical ganglia. When
rats with denerva,ted pineals were kept
in constant darkness or light, there was
no longer a difference in hydroxyindole
0-methyltransferase activity in the pin-
eal (21). In blinded rats, continuous
darkness or light had no effect on hy-
droxyindole 0-methyltransferase activi-
ty, which suggests that the retina is
necessary for transmission of light mes-
sages to the pineal. Bilateral lesions of
the media1 forebrain bundle, which
contains noradrenergic and serotonergic
nerves, also abolished the effects of

Pineal cell

Fig. 1. The pineal cell, sympathetic nerve,
and melatonin synthesis. Abbreviations:
TROH, tryptophan hydroxylase; AAD,
aromatic amino acid decarboxylase; 5 HT,
serotonin; NA T, serotonin N-acetyltrans-
ferase; MAO, monoamine oxidase; AcHT,
N-acetylserotonin; HIAA, S-hydroxyin-
doleaceiic acid; HIOMT, hydroxyindole
0-methyltransferase: and NA. noradrena-
line.

environmental lighting on pineal hy-
droxyindole 0-methyltransferase (22).

In a series of experiments (23) it
was shown that information about en-
vironmental lighting reaches the rat
pineal as follaws: retina --, inferior ac-
cessory optic tract + medial forebrairf
bundle -+ medial terminal nucleus of
the accessory optic system + pregangli-
onic sympathetic tract in the spinal
,cord + superior cervical ganglia + post-
ganglionic sympathetic fibers + paren-
chymal cells of the pineal.

Circadian Rhythms of the Pineal

Soon after melatonin was discovered,
relatively large amounts of its precursor
serotonin were found in the pineal
(24); the serotonin was evenly dis-
tributed between the parenchymal cells
and the sympathetic neNe terminals
(25). Quay (26) then found a marked.
24-hour cycle in the serotonin content
of the rat ,pineal. Peak levels of sero;
tonin were reached at about midday
(Fig. 2). Soon after nightfall, there
was a rapid fall in serotonin content
(27), The day-night rhythm of sero-
tonin content in the pineal persisted
unchanged in continuous darkness, but
was abolished in rats that here kept in
continuous light (27) (Fig. 2). Reversal
of the lighting schedule (light kept on
during thi night and off during the
daytime) changed the pineal serotonin
rhythm in the pineal by 180" within 6
days (28). All of these experiments

indicated that the daily rhythm in
pineal serotonin is endogenous (circa-
dian) but is synchronized by environ-
mental lighting. Denervation of the
pineal by removal of the superior
cervical ganglia abolished the serotonin
rhythm (27). Interruption of the nerve
impulses from the central nervous sys-
tem to the superior cervical ganglia
and depletion of brain noradrenaline
and serotonin with reserpine (29) also
suppressed the pineal serotonin rhythm.
These observations indicated that the
circadian rhythm of serotonin is gen-
erated by sympathetic nerve terminals
innervating the pineal, presumably by
changes in the release of the neuro-
transmitter noradrenaline. The circa-
dian serotonin rhythm appeared to be
generated by a "clock" in the brain.

The circadian rhythms in serotonin
content in the rat pineal appear as
early as 6 days after birth (30). When
lights were left on during the night,
`the nocturnal decline in serotonin con-
tent was prevented in adult and new-
born animals. Lights left on during the
night prevented the fall. in serotonin
in blinded IZday-old rats. When the
head of the 12-day-old rat was covered
with a hood and the lights were on,
the serotonin content fell at night.
After the hooded rats were 27 days
old, additional lighting no longer pre-
vented the deeline of serotonin at night
(30). Thus, &vironmental lighting can
reach the pineal gland .by an extra-
retina1 pathway in the newborn but not
in the adult rat. Extraretinal pineal re-
sponses have also been found in birds
(II).

There is a marked circadian rhythm
in pineal ZV-acetyltransferase (32 )
(Fig. 2)) which is 180' out of phase
with that for serotonin. One hour after
the onset of darkness,`there is a 30- to
SO-fold rise in the enzyme activity. A
circadian rhythm in the melatonin con-
tent of the rat pineal has the same
phasing as that of N-acetyltransferase.
Like the serotonin rhythm, the N-ace-
tyltransferase rhythm is abolished by
denervating the sympathetk nerves t0
the pineal or by interrupting nerve im-
pulses from the brain (32). Bilateral
lesions in the suprachiasmatic nucleus
(present in the hypothalamus) abol-
jsh the circadian rhythm of N-ace-
tyltransfeiase in the pineal (33). This
suggests that a biological clock in the
brain sends fibers through, innervates,
or is localized in the suprachiasmatic
nucleus.

The observation that the circadian
rhythms in pineal serotonin and N-ace-


tyltransferase are abolished by cutting
sympathetic innervation to the pineal
indicated that there might be differ-
ences in the release of noradrenaline
from these nerves during the day and
at night. Brownstein and I (34) found
a 24-hour rhythm in the turnover of
noradrenaline in the sympathetic nerves
innervating the pineal. More nor-
adrenaline is utilized (presumably by
release from nerve terminals onto the
pineal cell) at night than during the
day. This rhythm in turnover persisted
in blinded rats and was abolished in
continuous light. This strongly sug-
gested that the circadian rhythm in
the pineal cell is generated by diurnal
release of the neurotransmitter nor-
adrenaline. The circadian rhythm in
sympathetic nerve activity is presum-
ably driven <by a biological clock aris-
ing in or transmitted via the supra-
chiasmatic nucleus in the hypothalamus.

C!ontrol of Piaeal Indoleamine

Metabolkn in Organ Culture

The sympathetic nerves in the pineal
contain both noradrenaline and sero-
,tonin, and it was. not certain whether
these compounds released from the
nerve terminals exert their effects on
pineal indoleamines. If, as appeared
likely, the neurotransmitter noradrena-
line stimulates the pineal cell, does it
ad on an a- or S-adrenergic receptor,
and is adenylate cyclase involved? Also,
which is the critical step in the syn-
thesis of melatonin from tryptophan?

The rat pineal in organ culture proved
`to be a productive experimental tool to
answer these questions. In such a system,
the effects of biogenic amines, adrener-
gic blocking agents, and cyclic AMP
(adenosine 3',5'-monophosphate) could
be examined directly, free of the com-
plexities in viva. In studies on pineal
organ culture ini*tiated independently
by Shein and collaborators (35) and
Klein et al. (36), [14Cltryptophan was
added to the culture media and the
formation of radioactive serotonin, 5-
hydroxyindoleacetic acid (the deamin-
atd metabolite of serotonin) , and mela-
tonjn was measured. The pathway of
synthesis of melatonin in pineal organ
culture was found to be the same as in
the innervated pineal (Fig. 1). The
ad&ion of cycloheximide, an inhibitor
of protein synthesis, to the incubation
media completely inhibited the forma-
tion of [W!]wrotonin and [Wlmela-
tonin from [W]tryptophan, which in-
dicates that synthesis of new enzyme

Fig. 2. Circadian
rhythms in pineal
N - acetyltransferase,
serotonin, and mela-
tonin.

protein was obligatory for the forma-
tion of melatonin (37).

The addition of I-noradrenaline to
the pineal culture caused a marked
increase in the formation of radioac-
tive melatonin from tryptophan during
2 days of incubation (35-37). The
stimulation of melatonin synthesis in
the pineal organ culture by I-noradren-
aline was prevented by the addition
of I-propanolol, a /3-adrenergic block-
ing agent (37~). a-Adrenergic blocking
agents had no effect on the increased
formation of melatonin in the presence
of I-noradrenaline. When added to or-
gan cultures, a variety of sympathomi-
metic amines such as Z-adrenaline,
dopamine, octopamine, and tyramine
also stimulated the formation of mela-
tonin from tryptophan (37). Serotonin
or melatonin was without effect.

Many actions of noradrenaline are
mediated by cyclic AMP. Evidence that
noradrenaline might be acting via cyclic
AMP in the pineal came from the obser-
vation that .the catecholamine stimulatks
adenylate cyclase activity in homoge-
nates of rat pineal (38). The addition
. of cyclic AMP to pineal organ culture,
however, was without effect in stimu-
lating melatonin synthesis. Dibutyryl
cyclic AMP, a compound that, unlike
cyclic AMP, is not metabolized by
phosphodiesterase, markedly stimulated
the {ormation of [14C]melatonin from
[14C]tryptophan (39). These studies in
organ culture indicated that noradrena-
line released from sympathetic nerves
increases melatonin synthesis by stimu-
lating a p-adrenergic receptor on the
membrane of the pineal cell. This
.stimulation results in activation of
adenylate cyclase inside the cell to
make cyclic AMP.

Klein and Berg (40) examined the
effect of I-noradrenaline stimulation of
hydroxyindole 0-methyltransferase and
N-acetyltransferase, the enzyme that
converts serotonin to N-acefylserotonin.
Noradrenaline caused only a small in-

crease in hydroxyindole O-methyltrans-
ferase activity-not enough I to account
for the large increase in melatonin
formation. On the other hand, the cate-
cholamine caused about a 20-fold in-
crease in N-acetyltransferase activity
and melatonin formation after incuba-
tion of pineal organ culture .for 18
hours. Dlbutyryl cyclic AMP also
markedly stimulated the N-acetyltrans-
ferase activity. The increase in N-acetyl-
transferase activity by either l-por-
adrenaline or dibutyryl cyclic AMP was
blocked by the addition of protein syn-
thesis inhibitors to the culture media.
These observations suggested that the
regulation of N-acetyltransferase syn-
thesis by the p-adrenergic receptor is
the critical step-in the control of pineal
melatonin synthesis.

Regulation uf Pineal Circadian
Rhythms

N-Acetyltransferase (33)) serotonin
(26, 27). N-acetylserotonin (4Z), and
melatonin (42) undergo marked circa-
dian rhythms (Fig. 2). In view of the
diurnal rhythms in turnover of nor-
adrenaline in sympathetic nerves inner-
vating the rat pineal, it appeared likely
that the circadian rhythms in the in-
doleamines and N-acetyltransferase are
generated by day-night changes in the
release of the neurotransmitter (34).
Such a mechanism of regulation of
pineal circadian rhythms was estab-
lished by a series of experiments in
vivo (43). After the onset of darkness
at 1800 hours, N-acetyltransferase un-
derwent `a 30- to SO-fold increase in
activity (Fig. 3). In the first hour after
the lights were turned off at 1800
hours, only a small increase in enzyme
activity was observed. Beginning at
1900 hours there was a sharp increase
in enzyme activity, reaching a maxi-
mum at 2200 hours (Fig. 3). When
lights were kept on after 1800 hours,


/ Lights on or
propanolol in dark

:'

3  ,,I,, I,,,,,,,,,, J
t     1500       1800       2100      2400  t     0300    0600
lsoproterenol         Clock` hours         Isoproterenol
                                                1

Light               Dark

Light

Fig. 3. Induction and suppression of serotonin N-acetyltransferase in the pineal.

however, there was no increase in N-
acetyltransferase activity. Rats kept in
darkness during the day from 1000 to
1600 hours also showed no daytime
elevation in enzyme activity. These
findings showed that both darkness and
the proper setting of an internal clock
are necessary for the nighttime increase
in N-acetyltransferase activity. Inter-
rupting sympathetic nerve impulses by
bilateral removal of the superior cer-
vical ganglia or preganglionic decen-
tralization completely prevented the
nocturnal elevation of N-acetyltrans-
ferase (32) (Fig. 3).

When I-propanolol, a P-adrenergic
blocking agent, was administered to
rats before the onset of darkness, the
nighttime rise of enzyme was also
blocked (43) (Fig. 3). Injection of an (Y-
adrenergic blocking agent did not pre-
vent the elevation. Reserpine, a drug that
depletes nerves of both catecholamines
and serotonin, also prevented the
elevation of N-acetyltransfeiase. p-
Cblorophenylalanine, a compound that
depletes nerves. of serotonin, had no
effect on the circadian rhythm of N-
acetyltransferase, which indicates that
noradrenaline but not serontonin is in-
volved in inducing the enzyme. Cyclo-
heximide injected immediately before
the onset of darkness blocked the rise of
N-acetyltransferase, whereas actinomy-
tin D, an inhibitor of RNA synthesis,
had no such effect. These experiments
clearly showed that noradrenaline re-
leased from sympathetic nerves stimu-
lated the /3-adrenergic receptor which
in turn initiated events that lead to the
synthesis of new N-acetyltransferase
molecules.

thetic nerves, and a rapid fall in enzyme
Administration of the catecholamine
I-isoproterenol during the daytime when
both N-acetyltransferase activity and
N-acetylserotonin content are low and
serotonin content is high results in ele-
vation of the enzyme (Fig. 3) (44)
and its product N-acetylserotonin (41),
and a fall of serotonin (45). The day-
time rise in N-acetyltransferase and
N-acetylserotonin and fall in serotonin
induced by isoproterenol are blocked
by prior administration of a p-adre-
nergic blocking agent. These observa-
tions indicate that circadian rhythms of
pineal indoles are generated by changes
in N-acetyltransferase activity which in
turn are controlled by the /3-adrenergic
receptor. Thus, the rise in N-acetyltrads-
ferase at night causes a fall in sero-
tonin and a rise in N-acetylserotonin,
the precursor of mela,tonin (Fig. 2).
The reverse occurs during the daytime.
When rats are exposed to light during
the night, there is a precipitous fall
in pineal N-acetyltransferase (43, 46)
(Figs. 2 and 3). Isoproterenol injected
before the rats are exposed to light
prevents the light-induced decrease in
N-acetyltransferase activity (43). When
rats are returned to darkness after 10
minutes of light, N-acetyltransferase
activity ,begins to rise immediately and-
attains its initial level af.ter 3 hours.
Exposure to light during the night
causes a rise in seroton'in content as
enzyme activity falls (45). Thus, main-
,tenance of N-acetyltransferase activity
requires continuous occupation of the
P-adrenergic receptor. A brief expo-
sure to light reduces or shuts off the
release of nonadrenaline from sympa-

activity ensues (Fig. 3). The action
of light on the retina is a stimulatory
event, yet it inhibits the release of
the neurotransmitter from sympathetic
nerves in .the pineal. This would indi-
cate that, somewhere in the brain be-
tween the retina and the superior cervi-
cal ganglia, a stimulatory signal is con-
verted to an inhibitory signal.

When the /3-adrenergic blocking drug
propanolol is injected into rats at
night, N-acetyltransferase activity is
decreased to less than 15 percent of ibs
initial value within 10 minutes (43),
a reduction similar to that caused by
exposure to light (Fig. 3). Cyclohexi-
mide, given at night at a dose that im-
mediately inhibits protein synthesis in
vivo, results in a gradual decrease in
N-acetyltransferase activity with a half-
life of about 60 minutes (43). The
rapid decrease in pineal N-acetyltrans-
ferase activity after light exposure or
blyckage of the /3-adrenergic receptor
(half-life of 5 minutes) and the slower
fall in enzyme activi.ty when protein
synthesis is inhibited indicate at least
two mechanisms for the degradation
of N-acetyltransferase in vivo. The rap-
id decrease might result from the con-
version of an active to an inactive
form of the enzyme or from a dis-
aggregation of subunits of the enzyme
molecule. The slower decrease after in-
hibition of protein synthesis probably
represents normal degradation of the
enzyme.

ods of time might indicate de novo
The character of the N-acetylmtrans-
ferase induction by catecholamines de-
pends on the previous exposure to light.
When rats are given isoproterenol after
exposure to light for 6 hours, .there is
no increase in N-acetyltransferase for
the first hour (Fig. 3). Then, between
1 and 3 hours after isoproterenol ad-
ministr&ion,  enzyme  activity  rises,
reaching a maximum after 2 hours
(43) (.Fig. 3). When rats are Kept in
darkness from 1800 hours to 2400
hours and then placed in light for 10
minutes, there is the expected rapid
decrease in N-acetyltransferase activity.
After a brief exposure to light follow-
ing a prolonged period of darkness,
isoproterenol injection causes an imme-
diate increase in N-acetyltransferase
activity without the l-hour lag period
(43) (Fig. 3). Cycloheximide, admin-
istered just before isoproterenol, bloJts
both the immediate and delayed in-
creases in enzyme activity after isopro-
terenol injection. The delayed elevation
of enzyme .activity in the absence of
/3-adrenergic stimulation for long peri-


synthesis of new enzyme molecules.
The immediate elevation of N-acdyl-
transferase after a prolonged period
of stimulation of the P-adrenergic re-
ceptor in darkness indicates an accumu-
lation of messenger RNA or precursors
necessary for the synthesis of N-acdyl-
transferase.

In pineal organ culture, -noradrena-
link causes a rapid increase in cyclic
AhP (47, 48) as well as N-acetyl-
transferase and melatonin (40). Theo-
phylline, a phosphodiesterase inhibitor
enhances the effect of noradrenaline on
pineal cyclic AMP, while /3-adrenergic
blocking agents prevent this effect.
Thus noradrenaline stimulates N-acetyl-
transferase and melatonin production
via fl-adrenergic receptors linked to
adenylate cyclase and cyclic AMP for-
mation.

The temporal relationships between
stimulation of the pineal P-adrenergic
receptor, cyclic AMP content, and
formation of N-acetyltransferase in
vivo were reported by Deguchi (49).
Injection of I-isoproterenol during the
daytime (arrow. A in Fig. 4) resulted
in a 1%fold increase in pineal cyclic
AMP content within 2 minutes which
was maintained for 10 minutes; cyclic
AMP content then fell Co baseline lev-
els 30 minutes after administration of
the catecholamine. One hour after iso-
proterenol injection Chere was no
change in N-acetyltransferase activity;
during this time cyclic AMP content
reached a peak and then returned to its
initial level. Thirty minutes after cyclic
AMP content returned to baseline,
N-acetyltransferase activity rose, reach-
ing a maximum 3 hours after injection
of isoproterenol, and then returned to
the low daytime value after 5 hours
(Fig. 4). I-Propanolol injected before
isoproterenol (arrow A in Fig. 4) pre-
vented the rise of both cyclic AMP
content and of N-acetyltransferase ac-
tivity. When propanolol was injected
60 minutes after isoproterenol at a
time when cyclic AMP content had
already reached a peak and then re-
turned to its initial level (arrow B in
Fig. 4), the increase in N-acetyltrans-
ferase was still prevented. The admin-
istration of cycloheximide before iso-
proterenol (arrow A in Fig. 4) did not
affect the rise and fall of cyclic AMP
content, but it prevented the elevation
of N-acetyltransferase activity. When
the protein synthesis inhibitor was given
after the rise and fall of cyclic AMP
(arrow B in Pig. 4), the formation of
N-acetyltransferase was still blocked. It
can be concluded that stimulation of

600-

WO-                                 \               - 500  s
                                    \
     4           I'        \                  t
                                                         a
                                      \
400-            I'           \
                                        \     -4co j
           I'             \                0
                                                        %
300 -           :                          -300 f
                                            \             =
             :                     \      5
200 -      I                     \

                                               \     -200
                      ,                         \       g
                     ,
                     ,                            \       z
                    /
100 -                0
                  0
            #'

                     I        ?        I
t                    2         3         4

                llmr (hours)
A           B

Fig. 4. Relation between cyclic AMP, N-acetyhransferase, and p-adrenergic receptors,
I-Isoproterenol hydrochloride (5 mg per kilogram of body weight) was given intra-
venously to rats. Cyclic AMP and N-acetyltransferase activity was measured at times
indicated. Arrows A and B are the times at which I-propanolol (20 mg/kg) was in-
jected intravenously. Vertical bars indicate standard error of the mean. [From Deguchi
(49), with permission of Aoademic Press1

the /3-adrenergic receptor activates the
formation of cyclic AMP via adenylate
cyclase. This in turn induces the syn-
thesis of new N-acetyltransferase mole-
cules and thus more melatonin. Cyclic
AMP may act on the translational
process in proteip synthesis, since cy:
clohexamide (but not actinomycin D)
interferes with the induction of N-ace-
tyltransferase. It also appears ,that there
is an additional /3-adrenergic receptor
that is not involved in cyclic AMP
formation but is concerned wit: the
formation of N-acetyltransferase.

a hypothesis for the changes in respon-
siveness of postsynaptic sites on the
pineal cell after a variety of manipula-
tions.

Supersensitivity and Subsensitivity
in Pineal Response

N-Acetyltransferase activity in the
pineal is low during the daytime (32,
43). Various drugs and physiological
manipulations were studied for their
capacity to indu$e this enzyme during
the daytime (44). Adrenaline, L-dopa,
noradrenaline, and isoproterenol (Fig.
3) caused a marked increase in pineal
N-acetyltransferase activity when in-
jected during the daytime. The /3-adre-
nergic blocking agent I-propanolol pre-
vented the elevation of pineal N-
acetyl~transferase activity caused by the
drugs or stress, while an a-adrenergic
blocking agent did not.

An important and puzzling cellular   To examine whether these com-
phenomenon is the increase in respon-  pounds induced N-acetyltransferase di-
siveness of the end organ after denerva-  rectly or acted by the release of nor-
tion, decentralization, or administra-  adrenaline from the nerves innervating
tion of drugs. Prior administration of  the pineal, the pineal was denervated
cocaine or denervation of sympathetic  by removal of the superior cervical
nerves results in a markedly enhanced  ganglia or by chemical sympathectomy
physiologic response to noradrenaline  with 6-hydroxydopamine (44). When
(50). This increased response is due  the pineal was denervated, administra-
mainly to a presynaptic event: the  tion of L-dopa increased N-acetyltrans-
capacity of cocaine to block reuptake  ferase activity 100-fold as compared to
of noradrenaline  into  sympathetic  20- to 30-fold when the gland was
nerves (Itl), a major mechanism of in-  innervated. Thus, denervation caused
activation of the neurotransmitter (52).  an increased responsiveness (supersen-
Denervation of sympathetic nerves also  sitivity) to catecholamines. Supersensi-
causes suPersensitivity by destroying  tivity of denervated pineal with respect
the catecholamine uptake mechanism  to N-acetyltransferase was also ob-
and also by affecting the postsynaptic  served after administration of isopro-
sites (50). Recent work has suggested  terenol (48) or insulin or after stress


Table 1. Supersensitivity and subsensitivity in cultured pin&&. Rats were doncrvatod by bi-
lateral removal of the superior cervical ganglia 7 daye befom they were killed. Isoprotemnol-
treated rats received the drug (2.0 me/k& 8, 16, and 24 hours before they were killed. Pineals
wcra cultured for 10 hours with indicated concentrations of isoprdtemnol, and N-aeetyl-
transferase activity was measured.

Clsooroterenol

IV-Acetyltransferase (units)

Dcncrvated

I-Isoprotemnol-
  treeted

5 x 10-O           330240         1330'210'           262 7+
2 x 104          68Okl6O         2190 -c 330*           70*24*
1 x IO-'          940f 80         1180 % 150           320&50*
I x 10-a          17202 180         1490 -c 170          1380% 110

o P < .Ot compared to intact rats at the aam isoprotercnol concentration.

(53). The increased formation of N-
acetyltransferase was prevented by /3-
adrenergic receptor blocking agents or
by inhibitors of protein synthesis. These
experiments demonstrated that stimula-
tion of the pineal /3-adrenergic receptors
results in enhanced synthesis of the pro-
tein N-acetyltransferase (superinduc-
tion) after nerve denervation.

Superinduction of N-acetyltransferase
appeared within 24 hours after removal
of the ganglia innervating the` pineal
(44). It had been reported that pineal
adenylate cyclase became more sensi-
tive to noradrenaline, but not until 4
weeks after denervation (54). Thus,
the appearance of adenylate cyclase
supersensitivity does not explain the
rapid appearance of superinduction of
N-acetyltransferase. The /3-adrenergic
agonist I-isoproterenal was used to
examine whether the changes in re-
sponsivity occurred on the postsynaptic
membrane. This compound, unlike
noradrenaline, is not taken up into
sympathetic nerve terminals (55); any
presynaptic effects are thus eliminated.
Injection of small amounts of isopro-'
terenol into rats resulted in tenfold
greater induction of N-acetyltransferase
in denervated pineals as compared to
normally innervated pineals (48). When
large doses of isoproterenol were in-
jected, the same maximum induction
was achieved in the intact and de-
nervated pineals (48, 56).

Innervated and denervated pineals
were then cultured in varying amounts
of isoproterenol (48, 56). In the pres-
ence of the catecholamine, N-acetyl-
transferase activity increased gradually
and reached a maximum after 6 to 10
hours in culture. Protein synthesis in-
hibitors  or Sadrenergic blocking
agents prevented this rise. Maximum
induction of N-acetyltransferase in de-
nervated pineals in organ culture
occurred with an isoproterenol concen-
tration of 5 x lO-%f (Table l), while

enzyme activity in the innervated ,pineal
increased slightly at this concentration.
Maximum induction in the innervated
pineal was reached at an isoproterenol
concentration of 1 x IO-sM (Table 1).
Thus, the ,maximum response in the de-
nervated pineal occurred at a dose of
&adrenergic agonist about 1 percent
of that in the innervated gland. The
maximum enzyme activity achieved,
however, was the same in both cases.

To examine whether the increase in

i
   a

-let

.

Fig. 5. Supersensitivity and subsensitivity
of pineal N-acetyltransferase. Reserpine
(1.5 &kg) was injected subcutaneously
into rats 24 hours before they were killed.
A group of reserpine-treated rata received
I-isoproterenol ( 1.5 mg/kg) 8 and 16
hours before they were killed. Rats were
denervatcd by bilateral removal of the
superior cervical ganglia 48 hours before
death. One group of denervated rats re-
ceived I-isoproterenol (0.5 mg/kg) 8, 16,
24, and 32 hours before they were killed.
One group of rats was exposed to light
for 7 days.

sensitivity of the pineal response was
due to absence of the neurotransmitter
or to the nerve itself, rats were treated
with reserpine to deplete the sympa-
thetic nerve terminal of noradrenaline
(56). Injection of a small amount of
isoproterenol during the daytime re-
sulted in a greater increase in pineal N-
acetyltransferase activity in res?fpine-
treated rats as compared to nontreated
animals (Fig. 5). Again, when higher
doses of isoproterenol were injected,
there was no dilference in the maxi-
mum increase of N-acetyltransferase
activity in normal and drug-treated ani-
mals. Pineal glands of rats with intact
nerves that were depleted of noradrena-
line by reserpine also showed a marked
sensitivity in organ culture. The super-
sensitivity evoked by reserpine treat-
ment developed less than 24 hours
after the drug was given. Light reduces
the release of noradrenaline from
nerve terminals innervating the pineal.
As predicted, cultured pineals of rats
exposed to continuous lighting were
many times more responsive to induc-
tion of N-acetyltransferase by isopro-
terenol than were pineals from rats
under diurnal lighting conditions (56)
(Fig. 5). Pineals were about ten times
more responsive to cntocholamines at
the end of the light period (1800
hours) than at the end of the dark
period, 0600 hours (57). The large
(30.fold) elevation in pioeal N-acetyl-
transferase at night could be explained
by the reduced release of noradrenaline
during the daytime causing supenensi-
tivity to the increased neurotransmittcr
discharged with the onset of darkness.
Exposure to long periods of light also
resulted in increased elevation of pineal
cyclic AMP content after noradrena-
line injection (47).

These findings raised the possibility
that supersensitivity after depletion of
noradrenaline by reserpine or denerva-
tion might be prevented by the admin-
istration of the catecholamine. When
isoproterenol was injected twice in
a 24hour period to denervated or
reserpine-treated rats, not only was the
superinduction of pineal N-acetyltrans-
ferase suppressed, but there was a
markedly reduced induction of the en-
zyme when the pineal was later re-
moved and exposed .to isoproterenol
in organ culture (56) (Fig. 5).

The induction of N-acetyltransferase
in organ culture by low concentrations
of isoproterenol was considerably re-
duced in the pineals of isoproterenol-
treated rats (Table 1). However, the


maximum response at high concentra-
tions of the catecholamine was about
the same in pineals in organ culture in
both the isoproterenol-treated and un-
treated rats. Thus, continuous exposure
of the /?-adrenergic receptor to its
agonist catecholamine results in a re-
duced responsiveness, or tolerance.    )
The effect of various doses of isopro-

terenol on stimulation of cyclic AMP
in the pineal was examined 3 days
after pineal denervation (56). Injection
of small doses of isoproterenol resulted
in more than twice the cyclic AMP in-
crease in the denervated gland as in
the innervated pineal. Large doses of
isoproterenol gave the same maximum
increase of cyclic AMP content in in-
nervated and denervated pineals. Es-
sentially the same cyclic AMP increases
were obtained in innervated and de-
nervated pineal glands in organ culture.
There was no difference between the
intact and denervated pineals with
regard to induction of N-acetyltrans-
ferase activity when the organs were
cultured in the presence of dibutyryl
cyclic AMP (56), a compound that
acts intracellularly. This indicated that
denervation supersensitivity is due to
changes proximal to the site of action
of cyclic AMP, presumably on the
;8'-adrenergic receptor on the outer cell
membrane.

as compared to pineals exposed to a
complement of neurotransmitter nor-
mally discharged from sympathetic
nerve terminals. If the number of
catecholamine molecules impinging on
the /?-adrenergic rece$or is in&eked,
the pineal cell ibecomes less responsive
to the catechojamine (Fig. 6). Thus,
larger amounts of catecholamine are re-
quired to produce the same increase
of N-acetyltransferase in these glands
as compared to pineals receiving the
normal amount of noradrenaline. This
phenomenon is essentially the same as
tolerance.

The shift in dose response in the
induction of N-acetyltransferase by
catecholamine in supersensitive and
slibsensitive pineals (Table 1) indi-
cates a change in the affinity of the
receptor for its agonist isoproterenol.
Such changes in affinity suggest ihat
the conformation of the /3-adrenergic
receptor has been changed by prior
exposure to a larger or lesser amount

Normal response

The increase in cyclic AMP in pine-
als from reserpine-treated rats was two
to three times greater than in those
from untreated rats (56). Treating rats
with the protein synthesis inhibitor
cycloheximide did not block the in-
creased cyclic AMP response in re-
serpinetreated pineals. This experiment
suggests/ but does not prove that new
synthesis of receptor or of the adenyl-
ate cyclase system is not involved in
the development of supersensitivity.

Supersensjtive response

Subsensitive response

These experiments indicate that the
responsiveness  of the postsynaptic
P-adrenergic receptor and the forma-
tion of the pineal hormone depend on
previous exposure of the receptor to
its neurotransmitter. When the nor-
adrenaline discharged from sympathetic
nerves is decreased or eliminated by
denervation, decentralization, reserpine,
or light, the responsiveness of the p-ad-
renergic receptor of the pineal to cate-
cholamines and N-acetyltransferase in-
duction and, ultimately, to melatonin
forination is greatly increased (Fig. 6).
The same magnitude of induction of N-
acetyltransferase could be achieved with
a much smaller amount of catechola-
mine in noradrenaline-depleted pineals

Nerve terminal     Pineal cell
Fig. 6. Possible mechanisms for the de-
velopment of supersensitivity and sub-
sensitivity. After exposure to a reduced
number o# catecbolamine molecules, there
is an inc+ase in the coupling of the ,%
adrenergici receptor on the outside of the
pineal ce.11~ membrane to adenylate cyclase.
on the in&r membrane. This leads to in-
creases in the formation of cyclic AMP
and synthesis of iV-acetyltransferase when
the p-adrenergic receptor interacts with
catecholamines. The reverse occurs when
the &adren,ergic receptor is exposed to
excessive amounts of catecholamines; ATP,
adenosine triphosphate.

of its agonist. It is also possible that
receptor sites become more available
when exposed to few agonist molecules
or less available when exposed to an
excess of molecules. Changes in sub-
and supersensitivity can occur relative-
ly rapidly, which makes changes in
conformation or availability of the p-
adrenergic receptor a likely possibility
and a parsimonious adaptive mech-
anism for responsive cells.

Supersensitivity and tolerance to
physiologic agents and drugs is not un-
common. Repeated administration of
narcotic drugs results in rapid develop-
ment of tolerance. On the basis of
enzyme studies, it has been proposed
that tolerance to opiates is caused by
reduced availablity of receptor sites
resulting from a continuous interaction
of receptor with the narcotic drugs
(58). In denervated muscle, an in-
crease in the number of acetylcholine
binding sites appears on the postsynap-
tic membrane (59), and the reverse oc-
curs with exposure to large amounts
of cholinergic agents (60). The number
of insulin receptors decreases in fat
cells of `obese rats with high plasma
insulin concentrations (61). Adenylate
cyclase in the adrenal cortex is more
responsive to  adrenocorticotrbphic
hormone in rats in which the hormone
has been previously depleted by hypo-
physectomy (62). In view of our find-
ings with .the pineal, changes in many
physiologic  and pharmacologic re-
sponses might be due to alterations in
conformation or availability of receptor
sites of responsive cells imposed by
absence or excess of physiologic agents
and drugs.

There are circadian rhythms in sero-
tonin, serotonin N-acetyltransferase, N-
acetfiserotonin, and melatonin in the
pineal which persist in continuous dark-
ness and are abruptly abolished by
exposure to light. These rhythms are
generated by diurnal changes in the
release of the neurotransmitter nor-
adrenaline from sympathetic nerve ter-
minals innervating the pineal. An in-
creased discharge of noradrenaline at
night stimulates the /3-adrenergic re-
ceptor, which causes increased syn-
thesis of serotonin N-acetyltransferase
molecules inside the cell ,by mediation
of an adenylate cyclase system. As the
activity  of N-acetyltransferase  rises
during the night, the concentration of


its substrate, serotoniq, falls, and that
of the product, N-acetylserotonin, rises.
Increased synthesis of the pineal hor-
mone melatonin then follows as a result
of 0-methylation of N-acetylserotonin
by hydroxyindole O-methyltransferase.
The responsiveness of the pineal p-ad-
renergic receptor and the consequent
synthesis of N-acetyltransferase change;
the receptor becomes supersensitive
after decreased exposure to the cate-
cholamines noradrenaline' and isopro-
terenol and subsensitive after increased
exposure to the catecholamines. The
circadian rhythm in pineal amines ap-
pears to arise from a biological clock
present in or near the suprachiasmatic
nucleus in the hypothalamus. This
clock in turn is modulated by inhibition
by environmental light.

I. R. J. Wurtman. J. Axehod, D. E. Kelly.
Th# Pineal (Academic Press, New York,
I%@.

2. 1. A. Kappers, Z. Zelliorsch. Mikrosk. Anat.
ss, 163 (1960).

3. B. Do& and E. Hcerd, J. Neurophysiol. 25,
405 (1962).

4. A. B. Lerner, J. D. Cam, Y. Takaha@d, T. II.
La, W. Moore. J. Am. Chem. Sot. 86, 2587
(1958); A. B. Lerner, J. D. Case, R. V.
HeInxelman, ibid. 81, 6084 (1959).
5. A. B. Lerner and. J. D. Case, 1. Invest.
Dermatol. 32, 211 (1959).
6. J. Axelrod, W. B. Quay, P. C. Baker, Ndrure
(Lend.) 208. 3% (1965); F. C. G. Van der
Vaardonk. Cum. Mod. Biol. 1, 175 (1967).
7. R. J. Wurtman. J. Axelrod, B. W. Chu.
Science 141, 277 (l%S); -, J. E. Fischer,
Endocrfnolcwy 75, 266 (1964).
8. F. Francinl, R. Collu. L. Martini, in Gibe
Foundation SymposIwn on The PIneal Gland,
G. B. W. Wolstenholme and J. Knight, Eds.
$Ch+Ch;lll Livlngstone, London, 1971) pp.

9. ;I J32$ and S. Sorrentino, Jr., in fbtd.,
   . . .

10.

11.

12.

13.

14.

IS.

16.

17.

18.

19.

20.

21.

22.

23.

24.
ii

27.

28.

29.

30.

31.

32.

33.

34.

0. Romtekiev, L. Krullch, S. M. McCann,
Endocrinology 92, 1339 (1973).
S. Gaston and hf. Menaker, Science 160, 1125
mw.
:9,;;. Reiter, .&mu. Rev. Physlol.  35,  305

\-- .-,.
W. Lovenberg.  E. Jequier, A. Sjwrdama,
scisnce 1s 5. 217 (1%7).
H. Weissbach, B.` G. .Redfleld, J. Axelrod,
Bfochlm. Btophys. Acra.43, 352 (1960).
J. Axelrod and I-I. Weissbach, 1. Bfol. Chem.
236 211 110611
-- -, - _ _  \ _ _ - _, .
J. AneIrod, P. D. MacLean, R. W. Albers,
Ii. Weissbach. in Rsalonel Neurochemfstw.
S. S. Kety &td J. B&es, Eds. (Pergamon;
Oxford, l%l), pp. 307-311; W. B. Quay,
Life Sci. 4, 983 (1965); D. Cardlnale and
J. M. Rosner. Endocrinology 69, 301 (1971).
L. B. Browman, 1. Exp. Zool. 75, 375 (1937);
R. J. Wurtman, W. Roth, M. D. Altschule. J.
J. Wurtman, Acra Endocrlnol. 36, 617 (l%l).
V. M. Fiske, K. Bryant, J. Putnam, Endocrf-
nolo8y 66, 489 (1960).
R. J. Wurtman, J. Axelrod, L. S. Phillips,
Schce 142, 1071 (1963); J. Axelrod, R. J.
Wurtman, S. H. Snyder, 1. Biol. Chem. 240,
949 (1965).
D. E. Wolfe, L. T. Potter, K. C. Richardson,
J. Axelrqd, Science 136, 440 (1962).
R. J. Wurtman. J. Axehod, J. E. Fischer.
ibid. 143. 1328 (19641.
J. Axelrod. S. H. Snyder, A. Helkr, R. Y.
Moore. Ibid. 154. 898 (1966).
R. Y.. Moore, A. Heber, R. K. Bhatnager,
R. J. Wurtman, J. Axelrod. Arch. Neural.
16, 208 (1968).
N. J. Giarman and M. Day, Blochem.
Pharmocol. 1, 235 (1959).
C. Clwman, Pros. Brain Rrs. 10. 423 (1965).
W. B. Quay, Cen. Comp. Endocrinol. 3, 473
(1963).
S. H. Snyder, M. Zweig, J. Axelrod, J. E.
Fischer, Proc. Natl. Acad. Sci. U.S.A. 53,
301 (1965).

S. H. Snyder. J. Axelrod. M. Zweig, 1.
Pharmacol. Exp. Ther: 156, 2% (1%7).
S. H. Snyder and J. Axelrod, Se&race 149,
542 (1%5). -

M. Zweig, S. H. Snyder, J. Axelrod, PrOc.
Natl. Acad. Sci. U.S.A. 66, 515 (1966).
D. C. Klein and J. L. Weller, Science 169,
1093 (1970).
- R. Y. Moore, Proc. Natl. Acad. Sci.
U.S.A.' 68, 3107 (1971).
R. Y. Moore, In The Newosciences, Third
Study Pro~am.i F. 0. Schmitt and F. G.
Worden, Bds. (MIT Press, Cambridge, Mass.,
1974), pp. 537-547.
M. Brownstein and J. Axelrod. Science 184.
163 (1974).

35. H. M. Shein, R. J. Wmtman. I. Axelrod.
Nature (Lond.J 213, 730 (1967); R. J. Wurt-
man, F. Larin, J. Axelrod, H. M. She& K.
Rosasco, ibid. 217, 953 (1968).
36. D. C. Klein, G. R. Berg, I. Weller, Science
168, 979 (1970).

37. J. Axelrod, H. M. Sheia, R. J. Wurtman,
hoc. Natl. Acad. Scl. U.S.A. 62, 544 (1969).
37a. R. J. Wurtman, H. M. Shein, F. Larln,
  J. Neurochem. 18, 1683 (1971).
38. B. Weiss and E. Costa, Science 156, 1750
  (1%7)7): 1. Phormacol. Exu. Ther. 161. 310
iiwej:

39. H. M. Shein and R. J. Wurtman, Sckncc
  166. 519 (1%9).

40. D. C. Klein and G. R. Berg, Adv. BTochem.
Pharmacol.~ 3, 241 (1970).
41. M. Brownstein, J. hf. Saavedra, J. Axehod,
Mol. Pharmacol. 9, 605 (1973).
42. H. J. Lynch, Life Sci. IO, 791 (1971); C. L.
  Ralph, D. Mull, H. J. Lynch, L. Hedhmd.
Endocrinology 89, 1361 (1971).
43. T. Deguchi and J. Axlrod, Proc. Nail. Acad.
  Sri. U.S.A. 69, 2541 (1972).
44. -, ibid., p. 2208.
45. H. Illnerova, Life Sci. 10, 955 (1971); M.
  Brownstein. R. W. Holz. J. Axelrod, J.
Phormacol. Exp. The?. 186, 109 (1973).
46. D. C. Klein and J. L. WoUer, Science 177.
  532 (1972).

41. B. Weiss and S. I. Strada. Adv. Cyclic Nucl.

  Res. 1, 357 (1972).
46. T. Deguchi and J. Axclrod, Mol. Pharmocol.
  9, 612 (1973).
49. T. Dcguchi. ibid., p. 164.
50. U. Trendelenburg,  Pharmacol.  Rev.  15,  225

  (1%3).
51. L. G. Whitby, G. Hcrttlng, J. Axelrod.
Nature (Land.) 187, 604 (1966).
52. G. Herttbtn and J.  Axelrod. ibid.  192.  172

(1961). -

53. H. J. Lynch, J. P. Eng, R. J. Wurtman.
Proc. Natl. Acod. Sri. U.S.A, 70, 1704 (1973).
54. B. Weiss, J. Phwmocol, Exp. Ther. 166, 146
  (1969).

55. G. Hertting. Biochem. Pharmacol. 15. 1119

(1964). -

56. T. Deguchi and J. Axelmd, Proc. Notl. Aad.
Scf. U.S.A. 70, 2411 (1973).
57. J. Rotnero and J. Axehod, Science lU, 1691
  (1974).                      .:

56. j. &elrod, ibid. 124, 263 (1956).
59. R. Miledi and L. T. Pottrr, N&we (Land.)
233, 594 (1971).

60. C. C. Chang, T. F. Chen, S.-T. Chuang. J.
Physiol. (Land.) 230, 613 (1973).
61. R. G. Kahn, D. M. NeviBe, J. Roth, 1. Dial.
  Chem. 246, 244 (1973).
62. G. Sayers and R. J. Berg, Schnce 179. 1330
  (1973).