- Email: [email protected]
Serotonin-immunoreactivity in the adrenal medulla: Distribution and response to pharmacological manipulation
0361-9230184 $3.00 + .OO
Serotonin-Immunoreactivity in the Adrenal Medulla: Distribution and Response to Pharmacological Manipulation M. A. ~OLZWARTH,’ Department and *Department
C. SAWETAWAN
AND hf. S. BRO~FIELD*
of Anatomical
Sciences, University of Illinois, Urbana, IL. 61801 und Functional Sciences, School of Veterinary Medicine Madison, WI 53706 University of Wisconsin,
of Structural
Received 27 June 1984 Spmtctnin-immtmorra~~iviry in thcl adrenal HOLZWARTH, M. A., C. SAWETAWAN AND M. S. BROWNFIELD. medulla: ~istribu~ioil and response to phurrnll~(~i~~~i~u~ rnunip~~iuri[~n. BRAIN RES BULL 13(2) 299-308, 1984.-We previously found serotonin in the adrenal medulla of the rat using immunocytochemical methods. Serotonin immunoreactivity was found in 75% of the medullary cells and by double staining techniques was found to be present in epinephrinecontaining cells. In order to better understand
the chemical and pharm~ological nature of the adrenomedullary serotonin containing cells, we undertook the present studies to characterize the cells’ response to a number of agents which have been used to assess biochemical relationships in other serotonin systems. The responsiveness of the serotonin-containing cells to these agents was determined by immunocytochemical methods directed against serotonin. Reserpine, a monoamine depleting agent, caused a significant reduction in the amount of serotonin immunostaining. Parachloroamphetamine (PCA), a specific serotonin releaser, administered in viva had a minima1 effect on the depletion of serotonin immunoreactivity. However, when applied in vifro. PCA always caused a dose dependent depletion; the effect of PCA in virro was blocked by pretreatment and co-incubation with serotonin-uptake inhibitor, fluoxitine. Exposure to parachlorophenylalanine (PCPA) which inhibits serotonin synthesis, resulted in a marked reduction of immunostaining of most cells. The small population of cells which still stained intensely after PCPA treatment was depleted by incubation with the specific releasing agent, parachloroamphetamine (PCA) in vitro. Restoration of control-like immunostaining after reserpine-depletion occurred with treatment with the serotonin precursor. L-trytophan, both in vivo and in vitro. From these results we infer the existence and regulation of adrenomedullary serotonin which is similar but not identical to that reported for serotonin neurons of the
CNS. Serotonin
5.~ydroxytryptamine
Immunocytochemistry
Adrenal medulla
Serotonin-immunoreactivity
-.-
the adrenal medulla, a modified sympathetic ganglion, has been thought to secrete primarily catecholamines and other vesicle-associated substances [6,271.I¢ly, however, numerous putative peptidergic transmitters have been identified in cells and nerve processes in the adrenal medulla. Included among these are somatostatin [8], vasoactive intestinal peptide (VIP) [ IO,1 I], substance P [9], neurotensin [16], and enk~phalin [21]. Although the specific functions of these putative peptide transmitters are not yet known, it is likely that they have either a paracrine or an endocrine function. In addition yet another monoaminergic neurotransmitter, serotonin, has now been identified in the adrenal medulla. Recently we have presented immunocytochemical evidence for the presence of significant amounts of immunoreactive serotonin in rat adrenal medullary cells [ 12,131.The serotonin immunostaining is found in approximately 75% of adrenal medullary cells and appears to be co-distributed with epinephrine [12,13].
The present studies were undertaken to better understand the significance and the physiological control mechanisms of the adrenomedullary serotonin system by determining their response to a number of pharmacological agents which have been used to study other serotonin systems. These studies were carried out in viva and/or in vitro in order to enable us to determine whether the immunostaining changes observed reflected a direct action on the serotonin cells of the adrenal or an indirect action by acting on another serotonin producing system.
TRADITIONALLY
‘Requests for reprints should be addressed to M. A. Holzwarth, Illinois, Urbana, IL 61801.
EXPERIMENTAL
PROCEDURES
Animals All (N=73) animals used in these experiments were male Sprague-Dawley rats (Holtzmann) 100-150 g which were housed two per cage in a 12: 12 light-dark schedule with food and water available ad lib. Rats were handled daily for &7
Department of Anatomical Sciences, 506 South Mathews, University of
299
HOL%WAKTH, days prior to the beginning of an experiment to reduce the possible stress-induced effect on the adrenals in response to injections and at the time of sacrifice. To evaluate the properties of the serotonin containing chromaffin cells, rats were injected with pharmacological agents which are known to affect serotonin content in other serotonergic neurons and/or incubated in vitro with these agents; control animals were injected with saline and/or incubated in buffer alone. In each experiment, treated and control adrenals were prepared in parallel and the serotonm immunorea~~vity of the adrenal medulla evaluated and compared.
The modifcation of the unlabelJed antibody enzyme immunocytochemi~al method was used to demonst~te the presence and the relative amount of serotonin-like immunoreactivity [22] as described previously [12,13]. The immunocytochemical specificity of the serotonin antiserum and optimal dilutions have been determined and reported previously [12,13]. In brief, the serotonin antiserum exhibited no cross reactivity with catecholamines and appeared specific to serotonin especially in the 6-hydroxytetrahydrobe~-carboii~e form which is characteristic of the indoleamine after exposure to paraformaldehyde [12, 13, 201. Furthermore, no staining was observed when the first antibody in the staining sequence is omitted [ 12,131. ChIoral hydrate anesthetized rats were perfused through the heart with cold 4% par~o~~dehyde in 0.1 M phosphate buffer pH 7.4; the adrenals were removed and postfixed for two hours. After washing with graded series of buffers containing O%, 5%, 15%, and 30% sucrose, adrenals were cut at 20 p using a cryostat. Tissues from in vitro experiments were fixed by immersion in the same fixative for 5 hours and processed as above. Free floating sections were imthe peroxidase-antiperoxidase processed using munocytochemical method and modi~c~ion f221. Endogenous peroxidases were inactivated with 1% H,02 in 0.01 M phosphate buffered saline. Sections were then incubated in serotonin antibody (rabbit; ImmunoNuclear Corp.) which had been preabsorbed with bovine serum albumen (I mg/ml). The antibody was used at previously determined optimal dilutions of 1:250,000 and 1:500,000, for 18-24 hr at 4°C with gentle agitation. This was followed by incubation in 1:50 anti-rabbit gamma globulin antiserum (Miles Labs.) and 1:250 rabbit peroxidase anti-peroxidase complex (PAP, Cappel Labs) for 40 minutes each at room temperature with agitation. Phosphate buffered saline (0.01 M) containing 0.2% T&on X-100 and 1% normal sheep serum was used to dihrte the immune reagents and for washing of the sections prior to incubation with the immune reagents. Tissues were then washed with 0.05 M phosphate buffer, pH 7.6, and the bound peroxidase visualized using diaminobenzidine tetrachloride (DAB; 15 mg%) and 0.01% hydrogen peroxide in phosphate buffer; all sections were exposed to the DAB substrate simultaneously in a specially constructed tissue chamber. After further washing, sections were mounted on slides. coverslipped and examined and photog~phed under brightfield and differential interference contrast illumination (Nomarski). The distribution and intensity of the immunostaining of experimental groups was compared with that of the control group.
SAWE’l‘AW.-IC
A31ul) BKoWNNE’ir:l ,i)
sources of serotonin and thereby tictermme the autonomy 04 the adrenal medullary ser~tonin-~~~ntain~~g cells, in some experiments adrenals of pretreated or control rats were iucubated in vitro and exposed to various pharmacologiml agents. The individual experiments are outlined in the results section. Rats were decapitated and the adrenals were removed. Adrenals were maintained in ice cold Krchs Henseleit bicarbonate buffer pH 7.4 containing 2% glucose and gassed with 95% 0, and 5% CO,. .AdrenaJs, kept in the cold buffer, were quickly cleaned and bisected and then placed in Ehrlenmeyer flasks containing 10 ml buffer which was continuously gassed and maintained a: ?7Y’. After a 30 minute preincubation in buffer to allow for equilibration of the tissue, the buffer was replaced with fresh buffer containing the pharmacological agent to be tested and incubated for 2--3 hr: at least one change of buffer was made during the ~ncubati~~n period. Tissues were then placed in cold fixative for 5 hr and processed for immunocytochemistr-y 3s outlined above. WC know from previous experiments that adrenals incubated in this fashion remain viable for at Jcitst 4 hour-s.
The following pha~a~ologi~~ll agents were used in these experiments: parachlorophenylalaninc methyl ester HC’I (PCPA) (Sigma). parachloroamphctamine (PCA) (Regis). rcserpine (Sigma) or Serpasil (Ciba). Fluoxitinc (I.illy,. pargylinr
(Sigma).
I.-tryptophan
(Sigma,
If serotonin is a significant secretory product of adrenitl medullary cells, then regulatory processes should be present in these cells. To test this reasoning. a series of experiments was carried out to characterize the response of the adrenal medullary serotonin cells to pharmacological agents which are reported to release, inhibit synthesis, prevent uptake and stimulate synthesis of serotonin. These experiments were first carried out with the administration of drugs in I+c~. Subsequently because the observed changes in serotonin manipulations immunoreactivity after pharmacological might be the result of enher direrr effect on the adrenal or a reflection of an effect occurring cl\cwhere, the experiments were repeated irt ~irt’rr.
The serotonin immunostaining of the control adrenals appeared as reported previously [ 12.131.More than 75% of the adrenal medullary chromaffin cells stained positively for serotonin immunoreactivity within the cytoplasm. These cells were clustered into rosettes of varying numbers of cells: unstained cells were also clustered (Figs. 1 and 2). Jmmunostaining was also observed in those chromaftin cells which accompany the splanchnic nerve through the capsule and cortex. However, no immunostaining was observed either within the splanchnic nerve. in the adrenal cortex or in adrenal cortical nerves. Adrenal glands which had been maintained it7 ~Yfrtt for up to 3 hr demonstrated little change in immunostaining compared with itt vivc> controls. Although slightly paler in staining intensity than in the it7 IYI*O controls. the distribution and overall appearance of the immunoreactive cells was similar.
it&~c.rs c!/‘(i ,~~tt7~t~i~7it7~,I~~~i)~l*~it7r: i frtfg In
order to separate the adrenal gland from other possible
Serotonin
immunoreactivity
was reduced by the adminis-
SEROTONIN
IN THE ADRENAL
MEDULLA
FIG. 1. Serotonin immunoreactivity in adrenal medullary cells of the rat. The darkly stained serotonin-immunoreactive cells are in clusters which are separated by areas of unstained medullary cells. The empty spaces (*) are the lumena of blood vessels which were cleared of blood during the perfusion of the fixative. No immunostaining was observed in the splanchnic or other adrenal nerves or in the cortex except within the trail of medullary cells which accompany the splanchnic nerve through the cortex. Seroronin antiserum was applied at a dilution of 1:500,000. x74.
FIG. 2. Serotonin immunoreactivity in the cytoplasm of adenomedullary cells of control rat as seen with differential interference contrast (Nomarski). Serotonin antiserum was applied at a dilution of l:SOO,OOO.x356.
of reserpine 5 mg!kg IP 18-20 hr previously when compared to the simultaneously processed control animals. In addition to the overall decreased staining intensity of the serotonin-containing cells, a more patchy and ragged appearance of the immunostained cells was observed after the single injection (Fig. 3A). In spite of obvious behavioral manifestations, the serotonin depletion after reserpine treatment was not as great as anticipated. Reserpine-induced monoamine depletion is mediated by interference with vesicular untake and storage mechanisms r31 - . and indeed. similar to our observations here with serotonin, its effectiveness in the depletion of adrenal medullary catecholamines is not as great as that observed within the CNS 241. Therefore, we reasoned that the incomplete depletion of adrenal medulla~ serotonin might be the result of a relatively slower turnover rate of the adrenal serotonin or the result of nonvesicular
storage (cytopiasmic) of the indolamine 141 or the consequence of peculiar adrenomedullary vesicle characteristics. To determine whether SIOWturnover rate was responsible for our reserpine results, we tested whether the serotoninimmunoreactivity could be further reduced by daily injections of reserpine for 3 consecutive days; however, further depletion of serotonin in the adrenal was not evident.
tration
Effect of Specific Serotonin Releasing Agent We next sought to determine whether the immunoreacof the adrenal was sensitive to the specific neuronal serotonin releasing agent, p~a~hloroamphetamine (PCA). The actions of this drug are complex and incompletely understood but may include serotonin release, inhibition of monoamine oxidase. decrease of tryptophan hytive serotonin
HOLZWARTH.
SAWETAWAN
AND
BROWNPlLl,I)
FIG. 3. Serotonin-immunoreactivity of the adrenal medulla is decreased after resetpine and is restored by expobrrrc IO ,crotonin precur\lrr. L-tryptophan. A. Treatment of rats with reserpine (5 mg/kg. IP 18 hr previously) which depletes stores of scrotonin and catecholamtncs. results in an overall reduction of the adrenal medullary serutonin-immunostaining. This suggests these serofonin-cont;rinine. cells xc rc\ponsivc to manipulation of the uptake-storage mechanisms of the amine containing granules. B. To determine if scrotonin-immunostaining could reflect serotonin synthesis. reserpine-depleted rats were preloaded with the serotonin precursor. Ltryptophan (II’. 100 mg/kg) 4 hr and pargyline (I00 mg/kg) I hr prior to sacrifice. Serotonin-immunoreactivity is increased over that of rcserpine depleted :uirenomedullary cells These results suggest that adrenomedullary cells are capable of serc)tonin-synthesis or uptake of newly syntheGred transmitter Ser~~tonin antiserum dilution l:SOO.MM. x 179.
droxylase activity and decrease in high affinity uptake of serotonin; all of which result in a decreased serotonin content [ 191. PCA was administered IP at a dose of 10 mg/kg and the rats sacrificed 2, 5, 20 or 48 hr later. Although this dose is effective in causing the significant depletion of CNS serotonin [26], it had a lesser effect on adrenal medullary serotonin immunostaining. Five, 20 and 48 hr after IO mg/kg PCA. immunostaining was reduced relative to controls but not as significantly as after a single injection of reserpine (Table I). When the dose was increased to 20 mg/kg and two injections were given at 20 and 2 hr prior to sacriftce. the intensity of immunostaining was not different from that observed 2 hr after a single injection or after the lower dose (Table I). As noted above, PCA was not as effective a depleter of serotonin of the adrenal medulla as observed in the CNS: this is possibly not surprising considering that for example
the transmitter uptake mechanisms qtpcar not to functiort similarly in these modified axonless ganglion cells as in other neurons [I. 14. 261. Although the adrenal medulla is able to take up serotonin. both 5.7 dihydroxytryptamine and 6-hydroxydopamine. which are scrotonergic and catecholaminergic neurotoxins respectively whose toxicity depcndh on specific uptake mechanisms. arc ineffective within the adrenal medulla 1I]. Slow rates of uptake of PCA by adrenomedullary cells in view might also be responsible for the relative ineffectiveness of PCA to deplete serotonin. WC therefore determined the effectiveness of PCA as a releaset of adrenomedullary serotonin under ;,I iitrcj conditions: adrenals from untreated rats were incubated with PCA (5 X IO to 2x IO “ M). A dose related decrease in serotonin immunorcactivity was observed in those adrenals incubated with the serotonin releaser at doses of I x IO ’ M and greater as compared with the control incubated adrenals t Fig. 4;
SEROTONIN
IN THE ADRENAL
MEDULLA
Table I). At high concentrations (10 and 20 mM) of PCA, nearly no immunoreactivity was observed. This suggested that serotonin was essentially all released. Since PCA probably acts in the serotonin system by competitive inhibition, perhaps at several sites, we tested the possibility that a specific serotonin uptake inhibitor might block PCA’s entry and hence its effect. We found that adrenals preincubated with fluoxitine (5x IO-” M) for 30 minutes followed by a two hour incubation with both fluoxitine and PCA (1 x 10ml M), exhibited control-like immunostaining (Table 1). Fluoxitine specifically inhibits serotonin uptake in the CNS [28], and also inhibits the action of PCA by the prevention of its uptake into the serotonergic cells [ 191. These results suggest that (1) the decreased immunostaining in the presence of PCA in vitro is due to a decreased amount of serotonin presumably the result of release {see PCPA experiments below); and (2) that adrenomedullary cells are capable of uptake of transmitter related specific substances. The in vitro incubation of adrenals with fluoxitine alone had no effect on the immunostaining (Table 1).
To determine if the serotonin content of the adrenal medulla could be altered by the inhibition of serotonin synthesis, parachlorophenylalanine (PCPA) was administered to rats at a dose of 300 mg/kg, IP at 72 and 48 hr prior to sacrifice. Treatment with PCPA, which acts by the specific inhibition of t~ptophan hydroxylase, the enzyme required for the initial hydroxylation of the precursor tryptophan [ 151, resulted in a marked reduction in the immunostaining of most of the serotonin cells of the medulla. However, some cells scattered throughout the medulla stained even more intensely than immunoreactive cells in the untreated animals. Other than this intense immunostaining after PCPA treatment, these cells did not exhibit any obvious differentiating characteristics. To determine if the increased immunostaining was the result of drug-related cytotoxicity we examined 2 p plastic embedded sections of adrenals of control and treated rats; we found no abnormal cells. Further evidence for the viability of these intensely immunostained cells remaining after PCPA treatment is presented in experiments detailed below. In brief, when SPA-pretreated adrenals were exposed to the specific serotonin releasing drug, parachloroamphetamine (PCA), in vitro, the serotonin immunoreactivity was greatly reduced or absent. This indicates that the remaining PCPA-insensitive immunoreactive cells were capable of PCA uptake and serotonin release. As alluded to above, we tested the effect of the serotonin releasing drug, PCA, on PCPA-pretreated and control adrenals in an in vitro experiment. The pattern of immunostaining in the adrenals of PCPA-pretreated animals followed by 2-3 hr of incubation was essentially comparable to that of adrenals from rats which were PCPA treated and maintained in vivo; the marked overall reduction of immunostaining as well as the persistence of the darkly staining apparent PCPA-resistant cells were retained when adrenals were incubated in vitro. Interestingly, a number of the PCPA resistant cells exhibited long immunostained processes in vitro (see Fig. 5A). This was not seen in vivo. When PCPA-pretreated adrenals were incubated with PCA (1 or 2x 10m2M) (Fig. 5B), serotonin immunoreactivity was greatly reduced (1 x lo-’ M) or totally abolished (2x IO-’ M). The responsiveness of the darkly staining-SPA-resistant cells to PCA presumably demonstrates a pharmacological sensitivity which is similar
FIG. 4. Serotonin-immunoreactivity is reduced after in vitro exposure to specific serotonin releasing drug, parachloroamphetamine (PCA). Adrenals from untreated rats which were incubated with PCA (1 x 10ml M) for 2 hr demonst~ted a marked reduction of immunostaining as compared with simul~aeously incubated untreated control adrenals. At higher concentrations of PCA (1 x lo-’ and 2x 10eL M), concentrations which are reported to inhibit CNS tryptophan hydroxylase activity [18], nearly no immunoreactivity was observed. Serotonin antiserum dilution 1:500,000. x 184.
to the other serotonin containing cells. Since the tryptophan hydroxylase was already blocked, the PCA effect here is presumably due to its releasing action.
Effect of Serotonin Precursor In order to determine if immunost~ning in the adrenal medulla could also reflect serotonin synthesis, rats were first treated with reset-pine (5 mg/kg) for 18 hr to reduce endogenous serotonin stores and then injected with L-tryptophan (100 mg/kg, IP) at 4 hr and with pargyline 1 hr prior to sac&ice; saline-injected control rats were treated with pargyline alone. In other serotonin systems, preloading the animals with the serotonin precursor L-tryptophan, results in increased serotonin synthesis, while the addition of pargyline,
304
HOLZWARTH. TABLE
SAWETAWAN
AND BKOWNF11;1,1)
1
THE EFFECT OF PCA ON SEROTONLN-IMMUNOREACTIVITY
-
Treatment In Vi\Yj control
PCA: 10 mg/kg, 2 hr previously 5 or 20 or 48 hr previously 20 mgikg, 20 and 2 hr previously 2 hr previously In Vitro 1. control PCA: 5x 1O-5 M 1x 1O-4M IxIO-~ M 1x10_‘M 2x 1O-2M 2. control Fluoxitine: 5 x 10 ~’ M PCA: IxIO-~M PCA: 1 x lo-” M plus Fluoxitine: 5 x lo--’ M The effect of parachloroamphetamine (PCA), a specific serotonin releasing drug. on serotonin immunostaining of the adrenal medulla was much greater in Gtro than in CWI. Several doses of PCA were administered in t,ivo at various times prior to sacrifice or applied by in tfifro incubation for 2 hr. The effect of Fluoxitine on the depleting action 01 PCA was determined in l*irro in order to determine if the action of PCA could be blocked by a serotonin uptake inhibitor. Immunostained tissues were scored for relative intensity of staining from + t- + + to 0. Controls were run simultaneously.
a monoamine oxidase inhibitor, prevents the breakdown of the newly synthesized and released transmitter (2,261. When compared with controls, the reserpine-treated rats showed marked reduction in adrenomedullary serotonin immunostaining (as in the above experiment) while the reserpine plus L-tryptophan and pargyline treated rats resulted in immunostaining increased over the reserpine-pretreatment but still less than the controls (Fig. 3B). When this experiment was repeated using 200 mg/kg L-tryptophan, the increase in serotonin immunostaining in the previously depleted adrenal medulla was more pronounced than at the lower dose. Control rats exposed to the serotonin precursor failed to exhibit noticable increased immunostaining. We presume this was because our method is not very sensitive to further increases of such relatively high serotonin concentrations. In the partially depleted system, however, substrate enhancement of synthesis was easily demonstrated. The increased immunostaining observed after L-tryptophan pretreatment could be the result either of adrenomedullary serotonin synthesis or uptake of serotonin which had been synthesized elsewhere (e.g., gut). In order to determine if the serotonin is synthesized locally, we repeated the previous experiments, excepting that in these cases adrenals were exposed to the precursor L-tryptophan only in vitro. When the adrenals of reserpinedepleted rats were incubated for 2 hr in the presence or absence of L-tryptophan (Sx IO :Ior 5x lo-” M) and with pargyline (5~ IO--”M), exposure to the precursor resulted in an increase in immunostaining in the previously depleted adrenals (Fig. 6). The tissues exposed to the higher doses of L-tryptophan appeared to stain more intensely than those at the lower doses. These
results strongly suggest that dc, NOI‘Oserotonin curs within the adrenal medulla.
synthesis
OC-
DlSCUSSlON The results of these pharmacological-immunocytochemical studies provide evidence for the presence of a serotonin regulatory system in adrenai medullary cells of the rat. Evidence for the existence of serotonin in the adrenal medulla comes from four lines of evidence: (1) the specificity of the immunocytochemical method [12,13], (2) measurement of adrenal serotonin with radioenzymatic assay [26], (3) co-elution of serotonin standard with adrenal serotonin using reverse phase-ion pair high performance liquid chromatography with electrochemical detection by the method of Mayer and Shoup [17] (unpublished observations), and (4) in this communication, extensive pharmacological evidence. Altered serotonin immunostaining after the administration of specific serotoaergic agents strengthens the evidence for the specificity of the serotonin antibody and for the existence of a functional serotonin transmitter system within the adrenal medulla. Furthermore. the responses to serotonin drugs contribute to the identification of properties of these unique, modified sympathetic ganglion cells. While the adrenal medullary serotonin system possesses a number of characteristics which distinguish it from CNS serotonin neurons, this is not unique; other peripheral serotonin containing neurons also appear not to conform to all of the central serotonin characteristics [S]. As yet, there are no definitive explanations for why and how the central
and peripheral
serotonin
systems
differ.
SEROTONIN
IN THE ADRENAL
MEDULLA
305
FIG. 5. Serotonin-immunorea~~ivity of PCPA-resistant cells is further reduced by exposure to the specific serotonin-releaser (PCA) in vitro. (A) Serotonin-immunostaining in the adrenal medulla of rats which had been pretreated with serotonin synthesis inhibitor, PCPA (72 and 48 hr prior to sacrifice) followed by 3 hr incubation in buffer in vitro is greatly decreased compared with control. The darkly staining PCPA-resistant cells persist. Interestingly, a number of these cells exhibit immunostained processes (A). B. Serotonin-immunostaining is reduced when PCPA-pretreated adrenals are incubated with serotonin releasing agent, PCA (1 x lo-” M); when incubated with PCA at 2x lo-’ M, staining was totally abolished (not shown here). The responsiveness of the darkly stained PCPA-resistant cells to PCA presumably demonstrates a ph~acolo~cal sensitivity which is similar to that of the other adrenomedullary serotonin containing cells. Serotonin antiserum dilution 1:250.000. x 184.
HOLZWARTH,
SAWETAWAh
AND WKOWNPit:.ID
FIG. 6. Serotonin-immunostaining of reserpine-depkted adrenomedullary cells is restored after exposure to serotonin precursor in \irro. In order to determine whether the augmented serotonin staining observed after L-tryptophan preloading in viw. was the direct result of adrenomedullary cell serotonin synthesis. reserpine-depleted adrenals were incubated with L-tryptophan in r.ifro. A. Serotoninimmunostaining of reset-pine-depleted adrenomedullary cells incubated in vim in buffer for 2 hr is reduced compared with controls. 8. Serotonin-immunoreactivity is increased in reserpine-depleted adrenals after incubation with L-tryptophan (Sx IO 'M) for 2 hr. These results strongly suggest that dp now synthesis occurs within the adrenal medulla. Serotonin antiserum dilution 1:500.000. x 184.
SEROTONIN
IN THE ADRENAL
307
MEDULLA
Like CNS and enteric serotonergic neurons, the serotonin immunosta~ning of the adrenomedullary cells was at least partially depleted by reserpine and enhanced by the monoamine oxidase inhibitor, pargyline, indicating that the observable serotonin levels can be altered. It is noteworthy that both of these pharmacological agents are not specific to serotonin transmitter systems. The apparently reduced responsiveness to reserpine depletion observed in the adrenal medulla, as compared to brain, is unexplained. It is unlikely that it is the result of: (I) siower turnover rate because drug administration for 3 consecutive days did not further deplete serotonin stores; (2) specificity of the antibody to a compound which is serotonin-like but is not the same as the transmitter of the vesicle and which is released. More likely possibilities include: (1) serotonin is stored in a cytoplasmic as well as a vesicular pool [ 181 and since the reserpine acts at the level of the vesicle [3], a proportion of immunoreactive serotonin would remain unaffected by the drug; and (2) the transmitter-containing vesicles of the adrenal medulla differ from those of other neurons in that they are less sensitive to reserpine depleting mechanisms. However, until we know whether serotonin is stored in vesicles and is bound to serotonin binding protein(s) as it is in CNS neurons 1231, we cannot begin to explain this phenomenon. On the other hand PCPA, which is reported to be a specific inhibitor of tryptophan hydroxylase, the specific enzyme required for the synthesis of serotonin, substantially reduced the serotonin immunoreactivity of the adrenal medulla. These results corroborate a recent report of 70% decrease of radioenzymatically determined adrenal serotonin after PCPA treatment [26]. However, PCPA had no effect on the serotonin immunoreactivity in the enteric nervous system [5]. Although we have no definitive explanation for the remaining intensely stained adrenomedullary cells, we have evidence that they look normal and that they are responsive to the serotonin releasing drug, PCA. It is possible that because these particular cells merely store and release serotonin they are unaffected by the synthesis inhibitor. Another observation which marks differences between central and adrenomedullary serotonin neurons is the relative insensitivity of the adrenomedullary cells to PCA in viva; whereas in vifvo, PCA was easily demonstrated to have a dose-dependent effect. The differences in response of these cells to the specific releasing agent in vitro compared with in vivo are unexplained. but may be related to concentration differences, to increased access to the drug, or to the loss of their preganglionic innervation when in vitro, or less likely may be a simple consequence of being in vitro. Although specific transmitter uptake mechanisms of adrenal medullary cells may differ from those of other neurons, it appears that at least some uptake mechanisms are operational based on the results of the fluoxitine and PCA in vitro studies. On the other hand, the adrenomedullary cells are unresponsive to the neurotoxicity of 6-hydroxydopamine and 5,-7 dihydroxytryptamine, the specificity of both of which is related to specific uptake mechanisms [1,26], In view of a recent
report that chronic cultured adrenomedullary cells are sensitive to 6-hydroxydopamin~ presumably as a consequence of the growth of cell processes or of increased accessibility 1141, it would be of interest to determine if the above neurotoxins would be effective in our in vitro system. Results of the present studies indicate that the serotonin immunostaining of adrenomedullary cells can be increased by L-tryptophan in certain circumstances. The demonstration of augmented immunost~ning after exposure to the serotonin precursor following partial depletion of endogenous stores is interpreted to mean that adrenomedullary cells are capable of either serotonin synthesis or uptake of newly synthesized transmitter. By observing similar immunostaining when the adrenal was exposed to the precursor in vitro. as well as the observed sensitivity to PCPA, we conclude that the adrenomed~l~ cells are probably able to synthesize serotonin. Because preabsorption control studies demonstrated little crossreactivity with L-tryptophan or 5-hydroxytryptophan [ 12,131, it is uhlikely that the increased immunoreactivity reflects only uptake of the precursor. Interestingly, Costa and co-workers were unable to demonstrate serotonin synthesis in the enteric nervous system [S] while Gershon presents evidence for both tryptophan hydroxylase and for serotonin synthesis [7]. The appearance of immunostained nerve processes in tissues pharmacologically depleted and in vitro, was an unexpected observation and of considerable interest. Since the cells with processes often appeared different from the rest of the immunostained cells (larger and irregular in shape), it is possible that these are true ganglion cells and that for unknown reasons, are able to take up serotonin when incubated in vitro. That these processes are observed only in pharmacologically depleted adrenals may be because they are obscured by the abundance of reaction product in the undepleted adrenal. Another speculation is that the serotonin ~ont~ning ~hrorn~n cells, when removed from their normal physiological environment, are beginning to grow processes like medullary cells in culture, which tend toward reversion to typical sympathetic ganglion cells [24], or like chromaffin cells transplanted to the anterior chamber of the eye [25]. Further investigation is required to identify the cells which we have observed here. In concIusion, the serotonin immunoreactivity which has been demonstrated in epinephrine containing medullary cells of the rat adrenal can be depleted and augmented by pharmacological agents which have been shown to affect other serotonin containing neurons. As yet we have no definitive explanation why the adrenal medullary serotonin system demonstrates different sensitivities to serotonin pharmacological agents than the central serotonin neurons. ACKNOWLEDGEMENTS We are grateful to Dr. L. Barr for helpful discussions and criticism of the manuscript and to Dr. L. Van De Kar for providing a
number of pharmacological agents. This study was partially supported by NSF Grant PCM 8190756 and NIH Grant S-07RR7030.
REFERENCES 1. Aloe, L. and R. Levi-Montalcini,
Comparative studies on the effects elicited by pre- and postnatal injections of anti-NGF,
guanethidine, and 6-hydroxydopamine in chromaffin and ganglion cells of the adrenai meduIIa and carotid body in infant rats. fn: ~istf~~~e~~istry and Cell Bioiogy of A~tt~n~rn~~ Neurons, SIF Cells and Paranewons, edited by 0. Eranko, S. Soinila and H. Paivarinta. New York: Raven Press, 1980, pp. 221-226.
2. Boadle-Biber,
M. C. Biosynthesis of serotonin. In: Biology of edited by N. N. Osborne. New York: John Wiley & Sons Ltd., 1982, pp. 63-94. 3. Carlsson, A. In: Neurupsychopharmclcrtfogy. ~012, edited by E. Rothlin. Amsterdam: Elsevier Press, 1961. pp. 417-421. Serotonrrgic
Transmission,
308
HOl.%WAR’l-H.
4. Clementi, macological
F.
and
F.
on the catecholamines Rid
P.
Zocche.
Morphological
and
phar-
effects of reset-pine. given alone or after iproniazid.
17: 587-596.
of the adrenal
glands of the ;at. J Cc//
1963.
5. Costa, M., J. B. Fumess, A. C. Cuello, A. A. J. Verhofstad, H. W. J. Steinbusch and R. P. Elde. Neurons with 5-hydroxytryptamine-like immunoreactivity in the enteric nervous system: their visualization and reactions to drug treatment. h’c>urosc.ience 7: 35 l-363, 1982. 6. Euler. U. S. von and U. Hamberg. I-Noradrenaline in the suprarenal medulla. Narure 163: 642-643, 1949. 7. Gershon, M. D. Enteric serotonergic neurons. In: Biology td Srrotoncrgic Transmission. edited by N. N. Osborne. Ne\r York: John Wiley & Sons Ltd., l982.-pp. 363-399. 8. Hiikfelt. T.. L. G. Elfvin. R. Elde. M. Schultzbern. _. M. Goldstein and R. Luft. Occurrence of somatostatin-like immunoreactivity in some peripheral sympathetic noradrenergic neurons. Proc Nori Acad Sci USA 74: 3587-3591. 1977. 9. Hiikfelt, T., L. G. Elfvin. M. Schultzberg, M. Goldstein and G. Nilsson. On the occurrence of substance-P containing fibers in sympathetic ganglia: immunohistochemical evidence. Bruin Kv.\ 132: 2p-41. 1977. IO. H&felt, T., J. M. Lundberg, M. Schultzberg and J. Fahrenkrug. Immunohistochemical evidence for a local VIP-ergic neuron system in the adrenal gland of the rat. Acrcr Ph\.tkd St,and 113: 575-576, 1981. Il. Holzwatth, M. A. Localization of Vasoactive intestinal peptide (VIP)-like immunoreactivity in the adrenal cortex and medulla of the rat. J Arcron Ner\* Sysf. in press. 12. Holzwarth. M. A. and M. S. Brownlield. Serotoninimmunoreactivity in the adrenal medulla of the rat. .~I#( Neurosci Ahsrr 9: 388, 1983. 13. Holzwarth, M. A. and M. S. Brownfield. Serotonin coexists with epinephrine in rat adrenal medulla. (submitted) 14. Kirpekar, S. M., J. Nobiletti and J. M. Trifaro. Effect of bhydroxydopamine on bovine adrenal chromaffin cells in culture. Br J Pharmucol 79: 947-952. 1983. 15. Koe, B. K. and A. Weissman. pChlorophenylalanine: A specific depletor of brain serotonin. J Phcrrmucol Exp Thor 154: 499-516, 1966. 16. Lundberg, J. M., A. Rokaius, T. HGkfelt, S. Resell. M. Brown and M. Goldstein. Neurotensin-like immunoreactivity in prcganglionic sympathetic neurons and in the adrenal medulla of the cat. Acre Ph.wiol Sccrnd 114: ISF15.5. 1982.
SAWETAWA?,
AND
BROWN~lI;.l.L~
17. Mayer. G. S. and K. F.. Shoup. Sunultanrou\ rnult~ple clectrodc liquid chromatographic electrochermcal ahsay for catecholamines. indolamines and metaholites in brain tiz,sc.ic,ncc’ 5: 144% 1460. 1980 25. llnuicker. K.. B. Tschechne and I> I schechnc. t’ormation ot cholinergic synapses in adrenal chromaffin cells in ;mtsrior c\‘c chamber transplants. Brtrin Rc,\ lS2: 334-340. IY7X. X Van De Kar. I.. D.. (‘. W. WilkInson and W. F. Gsnong. Pharmacological evidence for a role of hr:un serotonin in the maintcnance of plasma renin activity in unanesthetized rat\. .I Plwr,,IC,C(I/ E.tp T/w 219: 8S-YO. 1981 27 Winklcr. H. and I!. Westhead. I‘he molecular orgamratton ol adrenal chromaftin granules. .\c,r/r< I\( i~~,lc~c~ 5: 180% 182.3. 1980. 2X. Wong. D. T.. J. S. Horng. I:. P. Bymaster. K. I.. Hauscr and B. B. Molloy. A selective inhibitor of xerotonin uptake: l.ill) 110140. i-(p-tri-fluoromethylphen~~l~y)-N-methyl-~-phenylpr~)pyl-amlne. Ii/i, Sci IS: 471 -479. 19’1
Serotonin-Immunoreactivity in the Adrenal Medulla: Distribution and Response to Pharmacological Manipulation M. A. ~OLZWARTH,’ Department and *Department
C. SAWETAWAN
AND hf. S. BRO~FIELD*
of Anatomical
Sciences, University of Illinois, Urbana, IL. 61801 und Functional Sciences, School of Veterinary Medicine Madison, WI 53706 University of Wisconsin,
of Structural
Received 27 June 1984 Spmtctnin-immtmorra~~iviry in thcl adrenal HOLZWARTH, M. A., C. SAWETAWAN AND M. S. BROWNFIELD. medulla: ~istribu~ioil and response to phurrnll~(~i~~~i~u~ rnunip~~iuri[~n. BRAIN RES BULL 13(2) 299-308, 1984.-We previously found serotonin in the adrenal medulla of the rat using immunocytochemical methods. Serotonin immunoreactivity was found in 75% of the medullary cells and by double staining techniques was found to be present in epinephrinecontaining cells. In order to better understand
the chemical and pharm~ological nature of the adrenomedullary serotonin containing cells, we undertook the present studies to characterize the cells’ response to a number of agents which have been used to assess biochemical relationships in other serotonin systems. The responsiveness of the serotonin-containing cells to these agents was determined by immunocytochemical methods directed against serotonin. Reserpine, a monoamine depleting agent, caused a significant reduction in the amount of serotonin immunostaining. Parachloroamphetamine (PCA), a specific serotonin releaser, administered in viva had a minima1 effect on the depletion of serotonin immunoreactivity. However, when applied in vifro. PCA always caused a dose dependent depletion; the effect of PCA in virro was blocked by pretreatment and co-incubation with serotonin-uptake inhibitor, fluoxitine. Exposure to parachlorophenylalanine (PCPA) which inhibits serotonin synthesis, resulted in a marked reduction of immunostaining of most cells. The small population of cells which still stained intensely after PCPA treatment was depleted by incubation with the specific releasing agent, parachloroamphetamine (PCA) in vitro. Restoration of control-like immunostaining after reserpine-depletion occurred with treatment with the serotonin precursor. L-trytophan, both in vivo and in vitro. From these results we infer the existence and regulation of adrenomedullary serotonin which is similar but not identical to that reported for serotonin neurons of the
CNS. Serotonin
5.~ydroxytryptamine
Immunocytochemistry
Adrenal medulla
Serotonin-immunoreactivity
-.-
the adrenal medulla, a modified sympathetic ganglion, has been thought to secrete primarily catecholamines and other vesicle-associated substances [6,271.I¢ly, however, numerous putative peptidergic transmitters have been identified in cells and nerve processes in the adrenal medulla. Included among these are somatostatin [8], vasoactive intestinal peptide (VIP) [ IO,1 I], substance P [9], neurotensin [16], and enk~phalin [21]. Although the specific functions of these putative peptide transmitters are not yet known, it is likely that they have either a paracrine or an endocrine function. In addition yet another monoaminergic neurotransmitter, serotonin, has now been identified in the adrenal medulla. Recently we have presented immunocytochemical evidence for the presence of significant amounts of immunoreactive serotonin in rat adrenal medullary cells [ 12,131.The serotonin immunostaining is found in approximately 75% of adrenal medullary cells and appears to be co-distributed with epinephrine [12,13].
The present studies were undertaken to better understand the significance and the physiological control mechanisms of the adrenomedullary serotonin system by determining their response to a number of pharmacological agents which have been used to study other serotonin systems. These studies were carried out in viva and/or in vitro in order to enable us to determine whether the immunostaining changes observed reflected a direct action on the serotonin cells of the adrenal or an indirect action by acting on another serotonin producing system.
TRADITIONALLY
‘Requests for reprints should be addressed to M. A. Holzwarth, Illinois, Urbana, IL 61801.
EXPERIMENTAL
PROCEDURES
Animals All (N=73) animals used in these experiments were male Sprague-Dawley rats (Holtzmann) 100-150 g which were housed two per cage in a 12: 12 light-dark schedule with food and water available ad lib. Rats were handled daily for &7
Department of Anatomical Sciences, 506 South Mathews, University of
299
HOL%WAKTH, days prior to the beginning of an experiment to reduce the possible stress-induced effect on the adrenals in response to injections and at the time of sacrifice. To evaluate the properties of the serotonin containing chromaffin cells, rats were injected with pharmacological agents which are known to affect serotonin content in other serotonergic neurons and/or incubated in vitro with these agents; control animals were injected with saline and/or incubated in buffer alone. In each experiment, treated and control adrenals were prepared in parallel and the serotonm immunorea~~vity of the adrenal medulla evaluated and compared.
The modifcation of the unlabelJed antibody enzyme immunocytochemi~al method was used to demonst~te the presence and the relative amount of serotonin-like immunoreactivity [22] as described previously [12,13]. The immunocytochemical specificity of the serotonin antiserum and optimal dilutions have been determined and reported previously [12,13]. In brief, the serotonin antiserum exhibited no cross reactivity with catecholamines and appeared specific to serotonin especially in the 6-hydroxytetrahydrobe~-carboii~e form which is characteristic of the indoleamine after exposure to paraformaldehyde [12, 13, 201. Furthermore, no staining was observed when the first antibody in the staining sequence is omitted [ 12,131. ChIoral hydrate anesthetized rats were perfused through the heart with cold 4% par~o~~dehyde in 0.1 M phosphate buffer pH 7.4; the adrenals were removed and postfixed for two hours. After washing with graded series of buffers containing O%, 5%, 15%, and 30% sucrose, adrenals were cut at 20 p using a cryostat. Tissues from in vitro experiments were fixed by immersion in the same fixative for 5 hours and processed as above. Free floating sections were imthe peroxidase-antiperoxidase processed using munocytochemical method and modi~c~ion f221. Endogenous peroxidases were inactivated with 1% H,02 in 0.01 M phosphate buffered saline. Sections were then incubated in serotonin antibody (rabbit; ImmunoNuclear Corp.) which had been preabsorbed with bovine serum albumen (I mg/ml). The antibody was used at previously determined optimal dilutions of 1:250,000 and 1:500,000, for 18-24 hr at 4°C with gentle agitation. This was followed by incubation in 1:50 anti-rabbit gamma globulin antiserum (Miles Labs.) and 1:250 rabbit peroxidase anti-peroxidase complex (PAP, Cappel Labs) for 40 minutes each at room temperature with agitation. Phosphate buffered saline (0.01 M) containing 0.2% T&on X-100 and 1% normal sheep serum was used to dihrte the immune reagents and for washing of the sections prior to incubation with the immune reagents. Tissues were then washed with 0.05 M phosphate buffer, pH 7.6, and the bound peroxidase visualized using diaminobenzidine tetrachloride (DAB; 15 mg%) and 0.01% hydrogen peroxide in phosphate buffer; all sections were exposed to the DAB substrate simultaneously in a specially constructed tissue chamber. After further washing, sections were mounted on slides. coverslipped and examined and photog~phed under brightfield and differential interference contrast illumination (Nomarski). The distribution and intensity of the immunostaining of experimental groups was compared with that of the control group.
SAWE’l‘AW.-IC
A31ul) BKoWNNE’ir:l ,i)
sources of serotonin and thereby tictermme the autonomy 04 the adrenal medullary ser~tonin-~~~ntain~~g cells, in some experiments adrenals of pretreated or control rats were iucubated in vitro and exposed to various pharmacologiml agents. The individual experiments are outlined in the results section. Rats were decapitated and the adrenals were removed. Adrenals were maintained in ice cold Krchs Henseleit bicarbonate buffer pH 7.4 containing 2% glucose and gassed with 95% 0, and 5% CO,. .AdrenaJs, kept in the cold buffer, were quickly cleaned and bisected and then placed in Ehrlenmeyer flasks containing 10 ml buffer which was continuously gassed and maintained a: ?7Y’. After a 30 minute preincubation in buffer to allow for equilibration of the tissue, the buffer was replaced with fresh buffer containing the pharmacological agent to be tested and incubated for 2--3 hr: at least one change of buffer was made during the ~ncubati~~n period. Tissues were then placed in cold fixative for 5 hr and processed for immunocytochemistr-y 3s outlined above. WC know from previous experiments that adrenals incubated in this fashion remain viable for at Jcitst 4 hour-s.
The following pha~a~ologi~~ll agents were used in these experiments: parachlorophenylalaninc methyl ester HC’I (PCPA) (Sigma). parachloroamphctamine (PCA) (Regis). rcserpine (Sigma) or Serpasil (Ciba). Fluoxitinc (I.illy,. pargylinr
(Sigma).
I.-tryptophan
(Sigma,
If serotonin is a significant secretory product of adrenitl medullary cells, then regulatory processes should be present in these cells. To test this reasoning. a series of experiments was carried out to characterize the response of the adrenal medullary serotonin cells to pharmacological agents which are reported to release, inhibit synthesis, prevent uptake and stimulate synthesis of serotonin. These experiments were first carried out with the administration of drugs in I+c~. Subsequently because the observed changes in serotonin manipulations immunoreactivity after pharmacological might be the result of enher direrr effect on the adrenal or a reflection of an effect occurring cl\cwhere, the experiments were repeated irt ~irt’rr.
The serotonin immunostaining of the control adrenals appeared as reported previously [ 12.131.More than 75% of the adrenal medullary chromaffin cells stained positively for serotonin immunoreactivity within the cytoplasm. These cells were clustered into rosettes of varying numbers of cells: unstained cells were also clustered (Figs. 1 and 2). Jmmunostaining was also observed in those chromaftin cells which accompany the splanchnic nerve through the capsule and cortex. However, no immunostaining was observed either within the splanchnic nerve. in the adrenal cortex or in adrenal cortical nerves. Adrenal glands which had been maintained it7 ~Yfrtt for up to 3 hr demonstrated little change in immunostaining compared with itt vivc> controls. Although slightly paler in staining intensity than in the it7 IYI*O controls. the distribution and overall appearance of the immunoreactive cells was similar.
it&~c.rs c!/‘(i ,~~tt7~t~i~7it7~,I~~~i)~l*~it7r: i frtfg In
order to separate the adrenal gland from other possible
Serotonin
immunoreactivity
was reduced by the adminis-
SEROTONIN
IN THE ADRENAL
MEDULLA
FIG. 1. Serotonin immunoreactivity in adrenal medullary cells of the rat. The darkly stained serotonin-immunoreactive cells are in clusters which are separated by areas of unstained medullary cells. The empty spaces (*) are the lumena of blood vessels which were cleared of blood during the perfusion of the fixative. No immunostaining was observed in the splanchnic or other adrenal nerves or in the cortex except within the trail of medullary cells which accompany the splanchnic nerve through the cortex. Seroronin antiserum was applied at a dilution of 1:500,000. x74.
FIG. 2. Serotonin immunoreactivity in the cytoplasm of adenomedullary cells of control rat as seen with differential interference contrast (Nomarski). Serotonin antiserum was applied at a dilution of l:SOO,OOO.x356.
of reserpine 5 mg!kg IP 18-20 hr previously when compared to the simultaneously processed control animals. In addition to the overall decreased staining intensity of the serotonin-containing cells, a more patchy and ragged appearance of the immunostained cells was observed after the single injection (Fig. 3A). In spite of obvious behavioral manifestations, the serotonin depletion after reserpine treatment was not as great as anticipated. Reserpine-induced monoamine depletion is mediated by interference with vesicular untake and storage mechanisms r31 - . and indeed. similar to our observations here with serotonin, its effectiveness in the depletion of adrenal medullary catecholamines is not as great as that observed within the CNS 241. Therefore, we reasoned that the incomplete depletion of adrenal medulla~ serotonin might be the result of a relatively slower turnover rate of the adrenal serotonin or the result of nonvesicular
storage (cytopiasmic) of the indolamine 141 or the consequence of peculiar adrenomedullary vesicle characteristics. To determine whether SIOWturnover rate was responsible for our reserpine results, we tested whether the serotoninimmunoreactivity could be further reduced by daily injections of reserpine for 3 consecutive days; however, further depletion of serotonin in the adrenal was not evident.
tration
Effect of Specific Serotonin Releasing Agent We next sought to determine whether the immunoreacof the adrenal was sensitive to the specific neuronal serotonin releasing agent, p~a~hloroamphetamine (PCA). The actions of this drug are complex and incompletely understood but may include serotonin release, inhibition of monoamine oxidase. decrease of tryptophan hytive serotonin
HOLZWARTH.
SAWETAWAN
AND
BROWNPlLl,I)
FIG. 3. Serotonin-immunoreactivity of the adrenal medulla is decreased after resetpine and is restored by expobrrrc IO ,crotonin precur\lrr. L-tryptophan. A. Treatment of rats with reserpine (5 mg/kg. IP 18 hr previously) which depletes stores of scrotonin and catecholamtncs. results in an overall reduction of the adrenal medullary serutonin-immunostaining. This suggests these serofonin-cont;rinine. cells xc rc\ponsivc to manipulation of the uptake-storage mechanisms of the amine containing granules. B. To determine if scrotonin-immunostaining could reflect serotonin synthesis. reserpine-depleted rats were preloaded with the serotonin precursor. Ltryptophan (II’. 100 mg/kg) 4 hr and pargyline (I00 mg/kg) I hr prior to sacrifice. Serotonin-immunoreactivity is increased over that of rcserpine depleted :uirenomedullary cells These results suggest that adrenomedullary cells are capable of serc)tonin-synthesis or uptake of newly syntheGred transmitter Ser~~tonin antiserum dilution l:SOO.MM. x 179.
droxylase activity and decrease in high affinity uptake of serotonin; all of which result in a decreased serotonin content [ 191. PCA was administered IP at a dose of 10 mg/kg and the rats sacrificed 2, 5, 20 or 48 hr later. Although this dose is effective in causing the significant depletion of CNS serotonin [26], it had a lesser effect on adrenal medullary serotonin immunostaining. Five, 20 and 48 hr after IO mg/kg PCA. immunostaining was reduced relative to controls but not as significantly as after a single injection of reserpine (Table I). When the dose was increased to 20 mg/kg and two injections were given at 20 and 2 hr prior to sacriftce. the intensity of immunostaining was not different from that observed 2 hr after a single injection or after the lower dose (Table I). As noted above, PCA was not as effective a depleter of serotonin of the adrenal medulla as observed in the CNS: this is possibly not surprising considering that for example
the transmitter uptake mechanisms qtpcar not to functiort similarly in these modified axonless ganglion cells as in other neurons [I. 14. 261. Although the adrenal medulla is able to take up serotonin. both 5.7 dihydroxytryptamine and 6-hydroxydopamine. which are scrotonergic and catecholaminergic neurotoxins respectively whose toxicity depcndh on specific uptake mechanisms. arc ineffective within the adrenal medulla 1I]. Slow rates of uptake of PCA by adrenomedullary cells in view might also be responsible for the relative ineffectiveness of PCA to deplete serotonin. WC therefore determined the effectiveness of PCA as a releaset of adrenomedullary serotonin under ;,I iitrcj conditions: adrenals from untreated rats were incubated with PCA (5 X IO to 2x IO “ M). A dose related decrease in serotonin immunorcactivity was observed in those adrenals incubated with the serotonin releaser at doses of I x IO ’ M and greater as compared with the control incubated adrenals t Fig. 4;
SEROTONIN
IN THE ADRENAL
MEDULLA
Table I). At high concentrations (10 and 20 mM) of PCA, nearly no immunoreactivity was observed. This suggested that serotonin was essentially all released. Since PCA probably acts in the serotonin system by competitive inhibition, perhaps at several sites, we tested the possibility that a specific serotonin uptake inhibitor might block PCA’s entry and hence its effect. We found that adrenals preincubated with fluoxitine (5x IO-” M) for 30 minutes followed by a two hour incubation with both fluoxitine and PCA (1 x 10ml M), exhibited control-like immunostaining (Table 1). Fluoxitine specifically inhibits serotonin uptake in the CNS [28], and also inhibits the action of PCA by the prevention of its uptake into the serotonergic cells [ 191. These results suggest that (1) the decreased immunostaining in the presence of PCA in vitro is due to a decreased amount of serotonin presumably the result of release {see PCPA experiments below); and (2) that adrenomedullary cells are capable of uptake of transmitter related specific substances. The in vitro incubation of adrenals with fluoxitine alone had no effect on the immunostaining (Table 1).
To determine if the serotonin content of the adrenal medulla could be altered by the inhibition of serotonin synthesis, parachlorophenylalanine (PCPA) was administered to rats at a dose of 300 mg/kg, IP at 72 and 48 hr prior to sacrifice. Treatment with PCPA, which acts by the specific inhibition of t~ptophan hydroxylase, the enzyme required for the initial hydroxylation of the precursor tryptophan [ 151, resulted in a marked reduction in the immunostaining of most of the serotonin cells of the medulla. However, some cells scattered throughout the medulla stained even more intensely than immunoreactive cells in the untreated animals. Other than this intense immunostaining after PCPA treatment, these cells did not exhibit any obvious differentiating characteristics. To determine if the increased immunostaining was the result of drug-related cytotoxicity we examined 2 p plastic embedded sections of adrenals of control and treated rats; we found no abnormal cells. Further evidence for the viability of these intensely immunostained cells remaining after PCPA treatment is presented in experiments detailed below. In brief, when SPA-pretreated adrenals were exposed to the specific serotonin releasing drug, parachloroamphetamine (PCA), in vitro, the serotonin immunoreactivity was greatly reduced or absent. This indicates that the remaining PCPA-insensitive immunoreactive cells were capable of PCA uptake and serotonin release. As alluded to above, we tested the effect of the serotonin releasing drug, PCA, on PCPA-pretreated and control adrenals in an in vitro experiment. The pattern of immunostaining in the adrenals of PCPA-pretreated animals followed by 2-3 hr of incubation was essentially comparable to that of adrenals from rats which were PCPA treated and maintained in vivo; the marked overall reduction of immunostaining as well as the persistence of the darkly staining apparent PCPA-resistant cells were retained when adrenals were incubated in vitro. Interestingly, a number of the PCPA resistant cells exhibited long immunostained processes in vitro (see Fig. 5A). This was not seen in vivo. When PCPA-pretreated adrenals were incubated with PCA (1 or 2x 10m2M) (Fig. 5B), serotonin immunoreactivity was greatly reduced (1 x lo-’ M) or totally abolished (2x IO-’ M). The responsiveness of the darkly staining-SPA-resistant cells to PCA presumably demonstrates a pharmacological sensitivity which is similar
FIG. 4. Serotonin-immunoreactivity is reduced after in vitro exposure to specific serotonin releasing drug, parachloroamphetamine (PCA). Adrenals from untreated rats which were incubated with PCA (1 x 10ml M) for 2 hr demonst~ted a marked reduction of immunostaining as compared with simul~aeously incubated untreated control adrenals. At higher concentrations of PCA (1 x lo-’ and 2x 10eL M), concentrations which are reported to inhibit CNS tryptophan hydroxylase activity [18], nearly no immunoreactivity was observed. Serotonin antiserum dilution 1:500,000. x 184.
to the other serotonin containing cells. Since the tryptophan hydroxylase was already blocked, the PCA effect here is presumably due to its releasing action.
Effect of Serotonin Precursor In order to determine if immunost~ning in the adrenal medulla could also reflect serotonin synthesis, rats were first treated with reset-pine (5 mg/kg) for 18 hr to reduce endogenous serotonin stores and then injected with L-tryptophan (100 mg/kg, IP) at 4 hr and with pargyline 1 hr prior to sac&ice; saline-injected control rats were treated with pargyline alone. In other serotonin systems, preloading the animals with the serotonin precursor L-tryptophan, results in increased serotonin synthesis, while the addition of pargyline,
304
HOLZWARTH. TABLE
SAWETAWAN
AND BKOWNF11;1,1)
1
THE EFFECT OF PCA ON SEROTONLN-IMMUNOREACTIVITY
-
Treatment In Vi\Yj control
PCA: 10 mg/kg, 2 hr previously 5 or 20 or 48 hr previously 20 mgikg, 20 and 2 hr previously 2 hr previously In Vitro 1. control PCA: 5x 1O-5 M 1x 1O-4M IxIO-~ M 1x10_‘M 2x 1O-2M 2. control Fluoxitine: 5 x 10 ~’ M PCA: IxIO-~M PCA: 1 x lo-” M plus Fluoxitine: 5 x lo--’ M The effect of parachloroamphetamine (PCA), a specific serotonin releasing drug. on serotonin immunostaining of the adrenal medulla was much greater in Gtro than in CWI. Several doses of PCA were administered in t,ivo at various times prior to sacrifice or applied by in tfifro incubation for 2 hr. The effect of Fluoxitine on the depleting action 01 PCA was determined in l*irro in order to determine if the action of PCA could be blocked by a serotonin uptake inhibitor. Immunostained tissues were scored for relative intensity of staining from + t- + + to 0. Controls were run simultaneously.
a monoamine oxidase inhibitor, prevents the breakdown of the newly synthesized and released transmitter (2,261. When compared with controls, the reserpine-treated rats showed marked reduction in adrenomedullary serotonin immunostaining (as in the above experiment) while the reserpine plus L-tryptophan and pargyline treated rats resulted in immunostaining increased over the reserpine-pretreatment but still less than the controls (Fig. 3B). When this experiment was repeated using 200 mg/kg L-tryptophan, the increase in serotonin immunostaining in the previously depleted adrenal medulla was more pronounced than at the lower dose. Control rats exposed to the serotonin precursor failed to exhibit noticable increased immunostaining. We presume this was because our method is not very sensitive to further increases of such relatively high serotonin concentrations. In the partially depleted system, however, substrate enhancement of synthesis was easily demonstrated. The increased immunostaining observed after L-tryptophan pretreatment could be the result either of adrenomedullary serotonin synthesis or uptake of serotonin which had been synthesized elsewhere (e.g., gut). In order to determine if the serotonin is synthesized locally, we repeated the previous experiments, excepting that in these cases adrenals were exposed to the precursor L-tryptophan only in vitro. When the adrenals of reserpinedepleted rats were incubated for 2 hr in the presence or absence of L-tryptophan (Sx IO :Ior 5x lo-” M) and with pargyline (5~ IO--”M), exposure to the precursor resulted in an increase in immunostaining in the previously depleted adrenals (Fig. 6). The tissues exposed to the higher doses of L-tryptophan appeared to stain more intensely than those at the lower doses. These
results strongly suggest that dc, NOI‘Oserotonin curs within the adrenal medulla.
synthesis
OC-
DlSCUSSlON The results of these pharmacological-immunocytochemical studies provide evidence for the presence of a serotonin regulatory system in adrenai medullary cells of the rat. Evidence for the existence of serotonin in the adrenal medulla comes from four lines of evidence: (1) the specificity of the immunocytochemical method [12,13], (2) measurement of adrenal serotonin with radioenzymatic assay [26], (3) co-elution of serotonin standard with adrenal serotonin using reverse phase-ion pair high performance liquid chromatography with electrochemical detection by the method of Mayer and Shoup [17] (unpublished observations), and (4) in this communication, extensive pharmacological evidence. Altered serotonin immunostaining after the administration of specific serotoaergic agents strengthens the evidence for the specificity of the serotonin antibody and for the existence of a functional serotonin transmitter system within the adrenal medulla. Furthermore. the responses to serotonin drugs contribute to the identification of properties of these unique, modified sympathetic ganglion cells. While the adrenal medullary serotonin system possesses a number of characteristics which distinguish it from CNS serotonin neurons, this is not unique; other peripheral serotonin containing neurons also appear not to conform to all of the central serotonin characteristics [S]. As yet, there are no definitive explanations for why and how the central
and peripheral
serotonin
systems
differ.
SEROTONIN
IN THE ADRENAL
MEDULLA
305
FIG. 5. Serotonin-immunorea~~ivity of PCPA-resistant cells is further reduced by exposure to the specific serotonin-releaser (PCA) in vitro. (A) Serotonin-immunostaining in the adrenal medulla of rats which had been pretreated with serotonin synthesis inhibitor, PCPA (72 and 48 hr prior to sacrifice) followed by 3 hr incubation in buffer in vitro is greatly decreased compared with control. The darkly staining PCPA-resistant cells persist. Interestingly, a number of these cells exhibit immunostained processes (A). B. Serotonin-immunostaining is reduced when PCPA-pretreated adrenals are incubated with serotonin releasing agent, PCA (1 x lo-” M); when incubated with PCA at 2x lo-’ M, staining was totally abolished (not shown here). The responsiveness of the darkly stained PCPA-resistant cells to PCA presumably demonstrates a ph~acolo~cal sensitivity which is similar to that of the other adrenomedullary serotonin containing cells. Serotonin antiserum dilution 1:250.000. x 184.
HOLZWARTH,
SAWETAWAh
AND WKOWNPit:.ID
FIG. 6. Serotonin-immunostaining of reserpine-depkted adrenomedullary cells is restored after exposure to serotonin precursor in \irro. In order to determine whether the augmented serotonin staining observed after L-tryptophan preloading in viw. was the direct result of adrenomedullary cell serotonin synthesis. reserpine-depleted adrenals were incubated with L-tryptophan in r.ifro. A. Serotoninimmunostaining of reset-pine-depleted adrenomedullary cells incubated in vim in buffer for 2 hr is reduced compared with controls. 8. Serotonin-immunoreactivity is increased in reserpine-depleted adrenals after incubation with L-tryptophan (Sx IO 'M) for 2 hr. These results strongly suggest that dp now synthesis occurs within the adrenal medulla. Serotonin antiserum dilution 1:500.000. x 184.
SEROTONIN
IN THE ADRENAL
307
MEDULLA
Like CNS and enteric serotonergic neurons, the serotonin immunosta~ning of the adrenomedullary cells was at least partially depleted by reserpine and enhanced by the monoamine oxidase inhibitor, pargyline, indicating that the observable serotonin levels can be altered. It is noteworthy that both of these pharmacological agents are not specific to serotonin transmitter systems. The apparently reduced responsiveness to reserpine depletion observed in the adrenal medulla, as compared to brain, is unexplained. It is unlikely that it is the result of: (I) siower turnover rate because drug administration for 3 consecutive days did not further deplete serotonin stores; (2) specificity of the antibody to a compound which is serotonin-like but is not the same as the transmitter of the vesicle and which is released. More likely possibilities include: (1) serotonin is stored in a cytoplasmic as well as a vesicular pool [ 181 and since the reserpine acts at the level of the vesicle [3], a proportion of immunoreactive serotonin would remain unaffected by the drug; and (2) the transmitter-containing vesicles of the adrenal medulla differ from those of other neurons in that they are less sensitive to reserpine depleting mechanisms. However, until we know whether serotonin is stored in vesicles and is bound to serotonin binding protein(s) as it is in CNS neurons 1231, we cannot begin to explain this phenomenon. On the other hand PCPA, which is reported to be a specific inhibitor of tryptophan hydroxylase, the specific enzyme required for the synthesis of serotonin, substantially reduced the serotonin immunoreactivity of the adrenal medulla. These results corroborate a recent report of 70% decrease of radioenzymatically determined adrenal serotonin after PCPA treatment [26]. However, PCPA had no effect on the serotonin immunoreactivity in the enteric nervous system [5]. Although we have no definitive explanation for the remaining intensely stained adrenomedullary cells, we have evidence that they look normal and that they are responsive to the serotonin releasing drug, PCA. It is possible that because these particular cells merely store and release serotonin they are unaffected by the synthesis inhibitor. Another observation which marks differences between central and adrenomedullary serotonin neurons is the relative insensitivity of the adrenomedullary cells to PCA in viva; whereas in vifvo, PCA was easily demonstrated to have a dose-dependent effect. The differences in response of these cells to the specific releasing agent in vitro compared with in vivo are unexplained. but may be related to concentration differences, to increased access to the drug, or to the loss of their preganglionic innervation when in vitro, or less likely may be a simple consequence of being in vitro. Although specific transmitter uptake mechanisms of adrenal medullary cells may differ from those of other neurons, it appears that at least some uptake mechanisms are operational based on the results of the fluoxitine and PCA in vitro studies. On the other hand, the adrenomedullary cells are unresponsive to the neurotoxicity of 6-hydroxydopamine and 5,-7 dihydroxytryptamine, the specificity of both of which is related to specific uptake mechanisms [1,26], In view of a recent
report that chronic cultured adrenomedullary cells are sensitive to 6-hydroxydopamin~ presumably as a consequence of the growth of cell processes or of increased accessibility 1141, it would be of interest to determine if the above neurotoxins would be effective in our in vitro system. Results of the present studies indicate that the serotonin immunostaining of adrenomedullary cells can be increased by L-tryptophan in certain circumstances. The demonstration of augmented immunost~ning after exposure to the serotonin precursor following partial depletion of endogenous stores is interpreted to mean that adrenomedullary cells are capable of either serotonin synthesis or uptake of newly synthesized transmitter. By observing similar immunostaining when the adrenal was exposed to the precursor in vitro. as well as the observed sensitivity to PCPA, we conclude that the adrenomed~l~ cells are probably able to synthesize serotonin. Because preabsorption control studies demonstrated little crossreactivity with L-tryptophan or 5-hydroxytryptophan [ 12,131, it is uhlikely that the increased immunoreactivity reflects only uptake of the precursor. Interestingly, Costa and co-workers were unable to demonstrate serotonin synthesis in the enteric nervous system [S] while Gershon presents evidence for both tryptophan hydroxylase and for serotonin synthesis [7]. The appearance of immunostained nerve processes in tissues pharmacologically depleted and in vitro, was an unexpected observation and of considerable interest. Since the cells with processes often appeared different from the rest of the immunostained cells (larger and irregular in shape), it is possible that these are true ganglion cells and that for unknown reasons, are able to take up serotonin when incubated in vitro. That these processes are observed only in pharmacologically depleted adrenals may be because they are obscured by the abundance of reaction product in the undepleted adrenal. Another speculation is that the serotonin ~ont~ning ~hrorn~n cells, when removed from their normal physiological environment, are beginning to grow processes like medullary cells in culture, which tend toward reversion to typical sympathetic ganglion cells [24], or like chromaffin cells transplanted to the anterior chamber of the eye [25]. Further investigation is required to identify the cells which we have observed here. In concIusion, the serotonin immunoreactivity which has been demonstrated in epinephrine containing medullary cells of the rat adrenal can be depleted and augmented by pharmacological agents which have been shown to affect other serotonin containing neurons. As yet we have no definitive explanation why the adrenal medullary serotonin system demonstrates different sensitivities to serotonin pharmacological agents than the central serotonin neurons. ACKNOWLEDGEMENTS We are grateful to Dr. L. Barr for helpful discussions and criticism of the manuscript and to Dr. L. Van De Kar for providing a
number of pharmacological agents. This study was partially supported by NSF Grant PCM 8190756 and NIH Grant S-07RR7030.
REFERENCES 1. Aloe, L. and R. Levi-Montalcini,
Comparative studies on the effects elicited by pre- and postnatal injections of anti-NGF,
guanethidine, and 6-hydroxydopamine in chromaffin and ganglion cells of the adrenai meduIIa and carotid body in infant rats. fn: ~istf~~~e~~istry and Cell Bioiogy of A~tt~n~rn~~ Neurons, SIF Cells and Paranewons, edited by 0. Eranko, S. Soinila and H. Paivarinta. New York: Raven Press, 1980, pp. 221-226.
2. Boadle-Biber,
M. C. Biosynthesis of serotonin. In: Biology of edited by N. N. Osborne. New York: John Wiley & Sons Ltd., 1982, pp. 63-94. 3. Carlsson, A. In: Neurupsychopharmclcrtfogy. ~012, edited by E. Rothlin. Amsterdam: Elsevier Press, 1961. pp. 417-421. Serotonrrgic
Transmission,
308
HOl.%WAR’l-H.
4. Clementi, macological
F.
and
F.
on the catecholamines Rid
P.
Zocche.
Morphological
and
phar-
effects of reset-pine. given alone or after iproniazid.
17: 587-596.
of the adrenal
glands of the ;at. J Cc//
1963.
5. Costa, M., J. B. Fumess, A. C. Cuello, A. A. J. Verhofstad, H. W. J. Steinbusch and R. P. Elde. Neurons with 5-hydroxytryptamine-like immunoreactivity in the enteric nervous system: their visualization and reactions to drug treatment. h’c>urosc.ience 7: 35 l-363, 1982. 6. Euler. U. S. von and U. Hamberg. I-Noradrenaline in the suprarenal medulla. Narure 163: 642-643, 1949. 7. Gershon, M. D. Enteric serotonergic neurons. In: Biology td Srrotoncrgic Transmission. edited by N. N. Osborne. Ne\r York: John Wiley & Sons Ltd., l982.-pp. 363-399. 8. Hiikfelt. T.. L. G. Elfvin. R. Elde. M. Schultzbern. _. M. Goldstein and R. Luft. Occurrence of somatostatin-like immunoreactivity in some peripheral sympathetic noradrenergic neurons. Proc Nori Acad Sci USA 74: 3587-3591. 1977. 9. Hiikfelt, T., L. G. Elfvin. M. Schultzberg, M. Goldstein and G. Nilsson. On the occurrence of substance-P containing fibers in sympathetic ganglia: immunohistochemical evidence. Bruin Kv.\ 132: 2p-41. 1977. IO. H&felt, T., J. M. Lundberg, M. Schultzberg and J. Fahrenkrug. Immunohistochemical evidence for a local VIP-ergic neuron system in the adrenal gland of the rat. Acrcr Ph\.tkd St,and 113: 575-576, 1981. Il. Holzwatth, M. A. Localization of Vasoactive intestinal peptide (VIP)-like immunoreactivity in the adrenal cortex and medulla of the rat. J Arcron Ner\* Sysf. in press. 12. Holzwarth. M. A. and M. S. Brownlield. Serotoninimmunoreactivity in the adrenal medulla of the rat. .~I#( Neurosci Ahsrr 9: 388, 1983. 13. Holzwarth, M. A. and M. S. Brownfield. Serotonin coexists with epinephrine in rat adrenal medulla. (submitted) 14. Kirpekar, S. M., J. Nobiletti and J. M. Trifaro. Effect of bhydroxydopamine on bovine adrenal chromaffin cells in culture. Br J Pharmucol 79: 947-952. 1983. 15. Koe, B. K. and A. Weissman. pChlorophenylalanine: A specific depletor of brain serotonin. J Phcrrmucol Exp Thor 154: 499-516, 1966. 16. Lundberg, J. M., A. Rokaius, T. HGkfelt, S. Resell. M. Brown and M. Goldstein. Neurotensin-like immunoreactivity in prcganglionic sympathetic neurons and in the adrenal medulla of the cat. Acre Ph.wiol Sccrnd 114: ISF15.5. 1982.
SAWETAWA?,
AND
BROWN~lI;.l.L~
17. Mayer. G. S. and K. F.. Shoup. Sunultanrou\ rnult~ple clectrodc liquid chromatographic electrochermcal ahsay for catecholamines. indolamines and metaholites in brain tiz