β-Endorphin levels in the cerebrospinal fluid of male talapoin monkeys in social groups related to dominance status and the luteinizing hormone response to naloxone

Neuroscience Printed

Vol. 18, No. 3,

in Great

pp. 651-658,

1986

0306-4522/86 $3.00 + 0.00 Pergamon Journals Ltd

Britain

0 1986IBRO

/?-ENDORPHIN LEVELS IN THE CEREBROSPINAL FLUID OF MALE TALAPOIN MONKEYS IN SOCIAL GROUPS RELATED TO DOMINANCE STATUS AND THE LUTEINIZING HORMONE RESPONSE TO NALOXONE N.

D. MARTENSZ, S. V. VELLUCCI, E. B. KEVERNEand J. HERBERT* Department

of Anatomy, University of Cambridge, Cambridge, U.K.

Abstrad-/?-Endorphin-like immunoreactivity was measured in the cerebrospinal fluid of 20 male talapoin monkeys living in mixed-sex social groups. It was shown that B-endorphin(,,,, was the major immu-

noreactive peptide; there was no evidence for high molecular weight precursors, or for either N-acetyl or C-shortened metabolites. Dominant males (those at the top of the social hierarchy) had lower levels of fi-endorphin than those of intermediate rank; subordinate males had higher levels than either of the other two ranks-about three times those measured in dominants. There were significant negative correlations between /3-endorphin in cerebrospinal fluid and both the amount of aggression given and sexual behaviour shown towards females. The response of the hypothalamo-pituitary system to opiate blockade was tested by giving the males naloxone in doses of 0.125, 0.25, 0.5, 1.0 and 5.0 mg/kg and assaying serum levels of luteinizing hormone 20min later. Dominant males released significant amounts of luteinizing hormone at doses of 0.25 and higher; there was no release in either intermediate or subordinate monkeys at any dose. These findings show that an animal’s rank in the social group in which it lives is strongly correlated with B-endorphin levels in the cerebrospinal fluid, and with changes in the neuroendocrine response to opiate blockade. Altered opiate neural activity may be responsible for the depressed levels of sexual behaviour and gonadal function observed in monkeys at the bottom of the hierarchy.

Although the detailed structure of primate social groups differs according to species, all share the common feature that individuals compete for limited resources, notably food and mates. Aggressive interactions reflect this competitiveness. If persistent aggressive interactions were necessary to determine an individual’s success or failure, there would be a high risk of repeated injury. Thus a number of mechanisms exist to reduce the need for aggression. One is to separate competitors (usually males) by forming relatively isolated family groups, a strategy adopted, for example, by some forest-living species. Other species, such as talapoin monkeys, which live in larger groups, form dominance hierarchies that act to maintain predetermined competitive status and so reduce aggression. 19pz6 A dominance hierarchy is recognisable by the direction of the aggressive interaction in the group; a dominant animal can displace or threaten a more subordinate member without retaliation. Subordination is a chronic condition which may persist for a large part of the lifespan of an adult monkey. It is characterised by a number of behavioural and endo-

*Author to whom correspondence should be addressed. Abbreviations: ACTH, adrenocorticotrophic hormone; CSF, cerebrospinal fluid; HPLC, high performance liquid chromatography; LH, luteinizing hormone; ODS, octadecylsilane column; TFA, trifluoroacetic acid.

crine states which develop in response to the threat of attack; a significant feature of these subordinate animals is their lowered reprdductive activity.16 This is reflected in the function of a subordinate’s neuroendocrine system vis-&is that of more dominant animals. Under laboratory conditions in which they are kept in mixed-sex groups, subordinate male talapoins display little or no sexual behaviour towards (or even interest in) the females, and their serum testosterone levels are often much lower than those of the most dominant male of the group.‘O However, subordinates show high levels of “visual monitoring” of dominants, presumably because the latter are a source of potential threat;20 their serum cortisol tends to be higher” and the relationship between the concentration of this hormone in the blood and in the cerebrospinal fluid (CSF) differs from that found in dominant males.” The neural mechanisms responsible for these coordinated behavioural and endocrine responses remain largely unknown. However, recent findings suggest that alterations in hypothalamic fl-endorphin activity might underlie at least some of them. Of the three opiate systems currently described in the brain, that containing /?-endorphin seems the best candidate for that regulating reproductive neuroendocrine function. Neurons synthesising b-endorphin lie in and around the hypothalamic arcuate nucleus,8’9 known as a site concerned with reproductive function. In-

651

N. D. MARTENSZel ol.

652

fusions of fi-endorphin, like some other opiates, inhibit both reproductive behaviour and testicular function.‘5.28.36 Increased levels have been reported in

the brain during reproductive suppression in other circumstances, such as the non-breeding season in hamsters,34 or pregnancy in rats.38 The experiments reported here investigate the activity of the intracerebral /I-endorphin-containing system in male monkeys of different social status. It is not feasible to measure /I-endorphin levels in the brains of talapoin monkeys. Though the evidence is not unanimous, it seems that fi-endorphin, like the related peptide adrenocorticotrophin, does not readily gain access to the cerebrospinal fluid (CSF) from the blood.‘~9~29~30~32 Therefore, levels in the CSF

are at least a guide to those in the extracellular fluid of the brain, and thus to the activity of the intracerebral /3-endorphin-containing systems. Furthermore, measuring CSF levels allows repeated estimates to be made on the same animal under different conditions. So far as we are aware, there have been no reports on the circumstances which can lead to altered concentrations of /I-endorphin in CSF under physiological conditions, though insulin-induced hypoglycaemia is reported to raise them.14 Post-translational processing of /I-endorphin seems likely to be an important mechanism regulating its biological activity. 23*24,4’ It is therefore necessary to establish the molecular form of the /I-endorphin-like peptides in the CSF since many metabolites will cross-react in a conventional radioimmunoassay. It is also helpful to have a more dynamic method of assessing alterations in opiate activity, with which to compare results from assaying levels in the CSF. In the study reported here, we have separated jI-endorphin from its metabolites in the CSF by high performance liquid chromatography (HPLC), and measured levels in males of differing social status and behaviour patterns. The effects of social status on the activity of the endogenous opioid peptide systems in terms of their inhibition of luteinizing hormone (LH) secretion were tested by measuring the LH response to acute pharmacological blockade using the opiate antagonist naloxone.s~‘2~27

MATERIALS

AND METHODS

Animals

Twenty intact male talapoin monkeys were studied. Seventeen were adult; the other three were juvenile and sexually immature (they had small testes and their serum testosterone levels were < 2 q/ml). All lived in one of six social groups under conditions previously descr&AtO Briefly, there were between 2 and 4 males per group. They had lived for a minimum of 6 months (and usualiy much longer) in two thirds of a large cage (3.5 x 1.5 x 1.7 m); the other third was occupied by 3 or 4 females, separated from the males by a solid metal partition except for mmin per day (4 days per week), during which the partition was removed and behavioural interactions scored by an automated procedure.‘o

Behaviour

Dominance hierarchies were constructed on the basis of the direction (not the amount) of aggression observed between the males. In all instances they were linear and stable over the time of this study. The top-ranking male in each group was the “dominant”; the lowest the “subordinate”; by definition, therefore, each group contained only one monkey of either category. Those in between were called “intermediates”, and their number depended upon the size of the group. Sexual behaviour was measured as the numbers of mounts made on the females. Aggression given was the number of threats or attacks made by each male; aggression received was the number of such behaviours received by a male which resulted in his withdrawing from the vicinity of the aggressor. Social behaviour was measured by the time spent huddling or grooming with other animals, and visual monitoring by the number of “glances” the males made to others. All behaviours are expressed as incidence per SOmin. Experimental design Experiment 1. Identity of fi-endorphin-rela!ed peptides in cerebrospinal fluid. Three large pools (2&30ml) of CSF

from the cisterna magna were obtained by combining consecutive samples (500 ~1) taken, respectively, from dominant, intermediate and subordinate males from 4 social groups. Each male was sampled 5-l 5 times, under ketamine anaesthesia (as described below). Each sample was frozen immediately after being taken and stored at -20’C. Immediately before assay the samples were thawed and pooled within each rank. Experiment 2. Correlations between cerebrospinal fluid /I-endorphin levels and social rank. Six groups of monkeys

were studied. Twice weekly, immediately after behavioural observations had been made, the males were tranquillised with ketamine (10 mg/kg), and 500 ~1 CSF taken from the cisterna magna. All the animals were highly accustomed to this procedure, which had been followed for several years. In this experiment, samples were taken from each male for 2-6 periods, each of 2-3 weeks. CSF samples were frozen immediately, and stored at -20°C. Immediately before assay, pools of CSF (2 ml) were made from each animal by combining 46 consecutive samples. This produced 36 pooled samples for each animal. Experiment 3. The effect of naloxone on luteinizing hormone levels in males of different social rank, Fifteen males

from 4 social groups were studied. Immediately after behavioural observations had been made, the males were tranquillised with ketamine (as above), and a venous blood sample taken. The males then received an intramuscular injection of naloxone HCl (Sigma Co., Poole, U.K.) and a second blood sample was taken after 20 min. The experiment was repeated at 2- to 4-day intervals, until each male had been given 0 (saline), 0.125, 0.25, 0.5, 1.0 and S.Omg/kg naloxone in a balanced design to reduce possible order of treatment effects. Blood was allowed to clot at room temperature for 1 h, stored overnight at 4°C and the serum removed after centrifugation and kept frozen at -20°C until assayed for LH. B-Endorphin Extraction from cerebrospinal jluid: experiments

1 and 2.

p-Endorphin was extracted from pooled samples by adsorption onto a small octadecylsilane column (ODS) (SepPak, Waters Associates. Cheshire. U.K.) as descrii bv Clement-Jones.’ The coiumn was equilibrated by washing with 3 ml 80% methanol in 0.1% trifluoroacetic acid (TFA) followed by 3 ml 0.1% TFA alone. After the CSF sample had been loaded, the column was washed with 2 ml 0. i% TFA. Adsorbed net&es were eluted with 2 ml 800/ methanol in 0.1% TFA.-Before the next sample was loaded, the column was re-equilibrated as described above. Each column was used for the extraction of 5 samples of CSF.

B-Endorphin in cerebrospinal fluid in talapoin monkeys B-endorphin-like

653

immunoreactivity

in the CSF of

The fraction containing /I-endorphin was dried at 47°C under reduced pressure and the residue dissolved in 600 ~1 of assay buffer (0.05 M phosphate buffer, pH 7.6, containing 0.25% human serum albumin and 0.001% thiomersalate) and duplicate aliquots (200~1) removed for radio-

male talapoins of each social rank co-elutecl with human p-endorphir+_,,, standard after gradient reversed-phase HPLC (Fig. 1). There was no evidence

immunoassay. Recovery -of iodinated Bendorphin

for either shortened C-terminal forms, or for Nacetvlated derivatives of B-endorohin. No significant

(25000cpm) was

84%. However, when 36fmol /I-endorphin was extracted from 2 ml 0.1% TFA, recoveries averaged 58%. High performance pooled samples of

liquid chromatography procedure cerebrospinal fluid (experiment

on

1). f?-Endorohin-like immunoreactivitv in CSF was senarated by reversed-phase HPLC using the Gilson gradient_HPLC svstem with two Gilson 302 numns (Anachem Ltd. Luton. C.K.) and a 5 pm ODS column (‘Novopak, Waters’ Associ: ates). 24 Elution was performed at 1.2ml/min using two buffers, one containing 0.1% TFA and 0.1% triethylamine, and the other 50% acetonitrile (HPLC grade, Rathburn Chemicals, Walkenburn, Scotland) in 0. I % TFA and 0.1%

triethylamine. The concentration of acetonitrile was increased from 25 to 32.5% over the first 5min, then by 0.1 %/min to reach 35% by 30 min, and finally to 50% by 35 min; it was kept at this level for 5 min before being

returned to the starting condition (25%). Fractions (0.6 ml) were collected between 7 and 27min and dried under reduced pressure. The residue was dissolved in buffer (as above) for RIA. The column was calibrated with 3OOfmol synthetic

human b-endorphin(,,,,, human p-endorphin(,,,,, and their acetylated metabolites (Universal Biologicals, London, U.K.), and human /I-lipotropin (NIADDK); they were oxidised with hydrogen peroxide. Recovery of all peptides exceeded 85 % The pooled samples (each 2&30 ml) from males of

different rank were passed through ODS columns in 2ml aliauots. senarated bv a wash with 1 ml 0.1% TFA. and the adsorbed peptides eluted as described above. After the solvent had been removed under reduced pressure, the

residue was re-dissolved in 200 ~150% acetic acid, and then oxidised for 1 h at room temperature by adding 60 rrl hydrogen peroxide (28% v/v). One hundred pl was Ljectkd into the HPLC columns. Radioimmunoassays. /I-Endorphin-like immunoreactivity was measured by radioimmunoassay using an antiserum which cross-reacted on an equimolar basis with human /I-endorphin+,,,, b-endorphinr,_,,, and their N-acetylated derivatives, and /?-lipotropin.24 The limit of the sensitivity of the assay was 3-4 fmol/tube, the intra-assay coefficient was 6.7%, and the interassay coefficient (assessed by routine inclusion of a pool of extracted human plasma: 18.1 fmol/ml) was 8.9%. Serum LH was measured by a heterologous radioimmunoassay used previously for both rhesus and talapoin monkevs.2,25The limit of sensitivitv was 0.09 ue LERl909-2 per tube, and between and wiihin assay ‘ciefficients of variation were 6.6 and 5.6%, respectively. Statistical procedures The data were analysed by two-way ANOVA (status x groups or treatment x status), dividing males according to their status into three categories (dominant, intermediate and subordinate); values for B-endorphin and LH were log-transformed; those for behaviour were con-

verted to squareroots. Where appropriate, posr hoc comparisons were tested with Duncan’s Multiple Range Test. Non-parametric Spearman correlations between p-endorphin levels and behavioural data were made on untransformed values. RESULTS

immunoreactive of /I -endorphin -like peptides in the cerebrospinal Jruid Experiment

The

1.

major

Identity

component

of

the

detectable

differences

were

observkd

in this

elution-pattern

between animals of the three social ranks, though the peak corresponding to human /I-endorphin was higher in subordinate males compared with the other two ranks (see b&w), Experiment 2. Levels of cerebrospinal j?-endorphin in males of dyerent ranks

@id

Mean levels of /I-endorphin in the CSF of male talapoins were lowest in the dominants, and highest in subordinates; values for animals of intermediate rank lay in between (Table 1). Analysis of variance confirmed that the effect of status was significant (F = 16.6; df= 2,17; P < 0.02). /I-Endorphin in dominants was significantly different from subordinates (Duncan’s test P < 0.05) but not from intermediates. There was no significant group x status interaction, and inspection of individual values showed that, in all six groups, levels of /?-endorphin in the CSF of subordinates was higher than that in dominants. There were only two (out of eight) instances in which an intermediate’s level of CSF B-endorphin was higher than that of the subordinate animal of its grouo. The concentration of B-endorohin in the CSF . -of the* three juveniles was similar to that of adults of equivalent rank. Correlations between cerebrospinal fluid /I-endorphin 1evels and behaviour The behaviour recorded from males of the three categories of social status in the six groups is shown in Fig. 2. Sexual behaviour, though not frequently

observed in the males of this experiment, was practically absent in both intermediates and subordinates,‘O which

ANOVA. more

precluded

satisfactory

Aggressive behaviour

frequently

by dominant

analysis

by

was shown much

males than by either

intermediates or subordinates (F = 3 1.1, df = 2; P < 0.01); the converse was observed for aggression received (F = 89.1; P < 0.02). Social behaviour was higher in both dominants and intermediates than in subordinates (F = 26.2; P < 0.01). Visual monitoring

Table 1. The concentration (mean -+ SEM) of b-endorphin in the cerebrospinal fluid of dominant, intermediate and subordinate male talapoin monkeys living in 6 social groups fl-Endorphin Dominant (n = 6) Intermediate (n = 8) Subordinate (n = 6) *P i 0.05 compared to subordinate n = number of animals.

(fmol/ml)

8.0 + 1.7* 14.7 f 1.2 23.6 + 7.0 monkeys.

N. D. MARTENSZet al.

654

Dominant a

e

b

I .,.................................,................................................ Intermediate 3+I

,

11

2-

4.

Subordinate 1111

I

___----

___---

____-----

_-I-

____----

___---

____---___--- __--

___--- --

- 35

-34 s L z s -33 3 c -32

....... .. . ... ............. ........ ...... ............ ..... ...... .......... ... ... ..

4 7

20 15 Retention time (min)

10

25

27

Fig. 1, HPLC separation of j-endorphin-like immunoreactivity in pooled samples of cerebrospinal’fluid from (top) dominant, (middle) intermediate and (bottom) subordinate male ta_JapoiFmonkeys. Etution

positions of synthetic human peptides are indicated by the arrows: (a) /?-$otropin, (b) fi-eadarphixq,_,,,, (c) acetyl-B-endorphin(,_,,,, (d) j?endorphin(,,,,, (e) acetyl-~-cndorphino_z,,. T& broken line represents the acetonitrile gradient and the dotted line represents the limit of sensitivity.

also showed rank-related differences, and was highest in subordinates (as previously described)” (F = 8.5; df = 2,17; P < 0.05). There were significant negative correlations (irrespective of rank) between levels of fi -endorphin in the CSF and both the amount of aggression given (Fig. 3) and sexual hehaviour displayed (the la&r analysis was limited to adult males) (Table 2). Conversely, there were correlations (though these were not significant) between fl-endorphin and aggression

received (Fig. 3) and social behaviour displayed. There was no significant positive correlation between visual monitoring and /II-endorphin. Experiment 3. Eflects of naloxone on luteinizing hormane in ma’es Ofdi@rent rank The effect of different doses of naloxone on increments in serum LH levels 20 min after the drug had heen given is shown in Fig. 4. The apparently higher levels in the dominant males than in either

Table 2. Correlation co&cients (Spearman) betweun cerebrospinal fluid fl-endorphin levels and b&&our in soeiaily living male talapoin mcmkeys of all m&s Aggr. recvd

Aggr.

behav.

given

Social behav.

Visual monitor

-0.57**

0.43

-0.67**

-0.40

0.29

Sex

/I-End.

**p < 0.01: IkesitiOns of bahavioural compomnta given in the text. Analysis of sexual bahaviour iimited to adult males (n = 17); others on all males (n = 20).

p-Endorphin in eerebrospinal fluid in talapoin monkeys

655

T

L

ILL-

S D I S D I S DIS DI IS Vis. mon/lO Aggro.gvnI2 Aggro.rcvdI2 Social Sexual Fig. 2. The incidence (mean f SEM) of sexual behaviour, aggression given and received, social behaviour D

and visual monitoring of (D) dominant, (I) intermediate and (S) subordinate male talapoin monkeys.

intermediate-ranking or subordinate males was not confirmed by ANOVA, because the number of animals in each group was so small. However, when the results from the lower two ranks were pooled, a clearly significant difference between them and the dominant animals was revealed (rank x treatment: F = 3.1;df = 11,84; P < 0.005). Naloxone did not

increase LH significantly above baseline in lowerranking animals, whereas all the doses of naloxone of 0.25 mg/kg and above did so in dominants (Duncan’s test; P < 0.05). But there was no evidence for a dose-related response. Similar results were obtained when absolute LH levels (rather than increments over baseline) were used for analysis (data not shown).

y . -0.66*+

25

y * 042r* 5.2 r i D42

14.9

ow

r.

20

.

.

. . . .

. . . .

.

.

. ..

\

3’

. 0

.

..

. 5

10 CSF $-morphin

15 iranksl

m

, 25

0

d

i

io CSF B-embrphm

.

.

is

io lranksl

Fig. 3. Correlation between CSF /I?-endorphin levels and the amount of aggression received (left) and aggression given (right) in socially living male talapoin monkeys of all ranks. *P < 0.05, Spearman rank correlations.

is

N.

D. MARTENSZet al. dominance rank in a monkey. /I-Endorphin in the CSF of talapoin monkeys of all ranks appears to bc principally /I-endorphin,, 3,) (the opiate active form). Subordinate monkeys have levels of b-endorphin in their CSF that are about three times those in dominants, and values for intermediate-ranking animals lie in between. Animals of different rank can also be distinguished by the reactivity to opiate blockade of the neuroendocrine system controlling LH secretion. Dominant males release significant amounts of LH in response to doses of naloxone as low as 0.25 mg/kg, whereas those as high as 5 mg/kg are ineffective in a group of intermediate and subordinate monkeys. These results show that the activity of a physiologically important peptide-containing cerebral system is different in males of different social rank in captive groups. The design of these experiments does not allow us to infer whether /I-endorphin levels reflect the dominance order or determine it, though the former seems more likely. Identity and source of /I-endorphin in cerebrospinal fluid

0~125

0.25

05

1.0

5.0

Naloxone hglkg) Fig. 4. The increment (mean f SEM) in serum LH concentrations 20min following various doses of naloxone in dominant, intermediate and subordinate male talapoin monkeys. The upper graph presents data from the three categories, whilst the lower graph shows the data from dominant animals and less dominant monkeys (intermediate and subordinate ranks). Top graph: +---0, dominant (n = 4); l ----a, intermediate (n = 7); C-0, subordinate (n = 4). Bottom graph: +---0, dominant (n = 4); C---C, less dominant (n = 11).

Since the pooling procedure used in this part of the study differed from that in the previous two experiments, the levels of /?-endorphin in the CSF were reanalysed, also pooling intermediate and subordinate ranks. Mean CSF /?-endorphin levels in domi-

was nant males (n = 6; x = 8.0 & 1.7 fmol/ml) significantly lower than that in males of lower rank (n = 14; x = 18.5 IfI 3.2 fmol/ml) (F = 4.4; df = 1; P < 0.05).

The experiments reported here correlate intracranial levels of the peptide fi-endorphin with social

Several studies on human CSF have identified the presence of /?-endorphin and related peptides.‘*~22~30~3’~33 Most have used gel filtration to separate members of the pro-opiocortin family; all seem to agree with our findings (using HPLC) that /?-endorphin(,_,,, is the major constituent. There seem to have been few deliberate attempts to detect N-acetylated or C-shortened forms, although these peptides are now recognised as important metabolites of /?-endorphin in the brain itself.4’ We were unable to show such metabolites in the CSF of talapoin monkeys, though small amounts of p-endorphin,, _r,) have been identified in the rat’s CSF.‘* Whether the metabolites occur in the brain of the talapoin is unknown; there seems to be considerable species variation-the hypothalamus of rats contains measurable amounts,” whereas that of another rodent, the hamster, has almost none (our unpublished results). Preliminary studies on the marmoset brain shown some lower molecular weight have /I-endorphin-like forms, which may represent these metabolites (our unpublished results). The conclusion from our study on the talapoin monkey is that nearly all the immunoreactivity we measured in the CSF is likely to be /?-endorphm,, ,,). Peptides derived from the anterior pituitary, of which ,!?-endorphin is one, cross the blood-brain barrier differentially.’ Whereas some of those of higher molecular weight, such as prolactin and LH, enter the CSF from the vascular compartment, other peptides such as adrenocorticotrophic hormone (ACTH) are excluded.’ Though there are some results pointing in the opposite direction, most of the available evidence suggests that /I-endorphin resembles ACTH. Thus, it seems likely that fl-endorphin in the CSF derives from that produced in the brain, principally by neurons lying in and around the hypo-

B-Endorphin

in cerebrospinal fluid in talapoin monkeys

thalamic arcuate nucleus. 3gSince CSF and the cerebral extracellular fluid are in equilibrium, and there is no effective barrier between them,3*‘3differences in CSF levels of /I-endorphin may reflect those in the brain’s extracellular space (assuming that the clearance of /3-endorphin from the CSF is similar in animals of different rank). It therefore seems that subordinate male talapoins have higher intracerebral extracellular levels of /I-endorphin than more dominant animals. Both this, and the behavioural and neuroendocrine profile of such animals, lead to the conclusion that this represents heightened /I-endorphinergic activity: this would be consistent with the low level of sexual behaviour shown by such animals, and their depressed gonadal function compared to dominant males. Though various inconclusive attempts have been made to show altered CSF /I-endorphin in various pathological states (such as schizophrenia) (e.g. Ref. 30), there has been no report of altered CSF /I-endorphin associated with a physiological condition such as the dominance heirarchy. Yet this is consonant with an increasing number of findings that show higher levels of b-endorphin in the brains of animals exposed to environmental conditions that inhibit reproduction, such as the non-breeding (see Introduction); the season, or pregnancy neuroendocrine effects of the social hierarchy may be a new example of an animal’s environment regulating reproductive activity through a possible action on the hypothalamic fl-endorphin system. Effects of opiate blockade by naloxone

There is an apparent inconsistency between the differential effects of opiate blockade with naloxone and postulated levels of activity of cerebral /I-endorphin in the monkeys studied here. Dominant males release LH in response to 0.25 mg/kg naloxone, a dose which compares with that given to humans.15 Lower-ranking male talapoins (that have higher CSF /I-endorphin) release little or no LH after being given naloxone, even in doses 10 times as great as in dominants. If the luteinizing hormone-releasing hormone pulse generator, and hence mean LH levels, were being inhibited by additional /I-endorphin in subordinates, would not this inhibition be released by opiate blockade? A similar paradox is presented by the photo-inhibited hamster, which also has low LH

651

levels, and higher hypothalamic /I-endorphin than reproductively active animals.“*” Photo-inhibited hamsters also do not release LH in response to naloxone.35 It is generally agreed that naloxone alters pituitary hormone output by an indirect action on the brain, rather than on the gland itself.37 Naloxone has a preferential, but by no means exclusive, action on mu (rather than delta or kappa) receptors;2’*40and it is therefore possible that many of the endocrine actions of naloxone (including the release of LH) are mediated through this class of receptor.” However, it is certainly not justifiable to conclude that LH release by naloxone represents specific antagonism of /.I-endorphin, rather than other opiates. Nevertheless, there seems to be a correlation developing between elevated intracranial /I-endorphin levels and naloxone insensitivity. A confident explanation cannot be offered, though persistent suppression of the luteinizing hormone-releasing hormone pulse generator by an input (such as /I-endorphin) may prevent its response to the relatively brief release following naloxone. Recent studies suggest that release of LH in response to naloxone can be altered by prior exposure to opiates.6 Though the exact interpretation of our results with naloxone is not yet clear, they reinforce the conclusion that chronic subordination is associated with altered opiate activity, and that this may be part of the neuroendocrine machinery of the reproductive suppression characteristic of such a state. Behavioural correlates of /I-endorphin in cerebrospinal fluid

Though it is not yet possible to pinpoint the behavioural mechanism responsible for altered levels of /I-endorphin, there was a significant correlation between the amount of aggression given and /I-endorphin levels in the CSF. The plausible suggestion that this behaviour may lead to increased /I-endorphin levels, which then depress reproductive behaviour, is one which remains to be tested by further experiments. Acknowledgements-Supported by a programme grant from the Medical Research Council. We thank Kathy Batty, Carlos de la Riva and Helen Shiers for expert technical help, and Jane Rowe11 for preparing this manuscript.

REFERENCES 1.

2. 3. 4. 5.

Beckford U., Herbert J., Jones M. T., Martensz N. D., Nicholson S. A., Gillham B. and Hamer J. D. (1985) Relationship between adrenocorticotrophin bioactivity in blood and cerebrospinal fluid of rhesus monkeys. J. Endocr. 104, 331-338. Bowman L. A., Dilley S. R. and Keveme E. B. (1978) Suppression of oestrogen-induced LH surges by social subordination in talapoin monkeys. Nature 275, 5658. Bradbury M. (1979) The Concept of a Blood-Brain Barrier. Wiley, Chichester. Cicero T. J. (1984) Opiate-mediated control of luteinizing hormone in the male: physiological implications. In Opioid Modulation of Endocrine Function (eds Delitala G. et al. ), pp. 211-222. Raven Press New York. Cicero T. J., Schainker B. A. and Meyer E. R. (1979) Endogenous opioids part&ate in the regulation of the hypothalamic-pituitary-luteinizing hormone axis and testosterone’s negative feedback control of luteinizing hormone. Endocrinology 104, 1286-1291.

N. D. MARTENS et al.

658

6. Cicero T. J., Owens D. P., Newman K. S., Schmoeker P. F. and Meyer E. R. (1984) Exogenous and endogenous opioid-induced enhancements of naloxone’s effects on serum luteinizing hormone levels in the male rat. Eur. J. Pharmur. 105, 273-284. I. Clement-Jones V., Lowry P. J., Rees L. H. and Besser G. M. (1980) Development of a specific extracted radioimmunoassay for methionine enkephalin in human plasma and cerebrospinal fluid. J. Endocr. 86, 231. 243. 8. de Kloet E. R., Palkovits M. and Mezey E. (1981) Opiocortin peptides: localisation, source and avenues of transport. Pharmac. Ther. 12, 321-351.

9. Dubey A. K., Herbert J., Martens2 N. D., Beckford U. and Jones M. T. (1983) Differential penetration of three anterior pituitary peptide hormones into the cerebrospinal fluid of rhesus monkeys. Life Sci. 32, 185771863. 10. Eberhart J. A., Keveme E. B. and Meller R. E. (1980) Social influences on plasma testosterone levels in male talapoin monkeys. Horm. Behav. 14, 247-266. 11. Eberhart J. A., Keverne E. B. and Meller R. E. (1983) Social influences on the circulating levels of cortisol and prolactin in male talapoin monkeys. Physiol. Behau. 30, 361-369. 12. Fabre-Nys C., Meller R. E. and Keveme E. B. (1982) Opiate antagonists stimulate affiliative behaviour in monkeys. Pharmac. Biochem. Behav. 16, 653659.

13. Fenstermacher J. D. and Patlak C. S. (1975) The exchange of material between cerebrospinal fluid and brain. In Fluid Environment ofthe Brain (eds Cserr H. F., Fenstermacher J. D. and Fencl V.), pp. 201-214. Academic Press, New York. 14. Fukata J., Nakai Y., Endo K. and Imura H. (1982) Hypoglycemia-induced elevation of immunoreactive /I-endorphin in cerebrospinal fluid in the cat. Brain Res. 246, 164-167. 15. Grossman A. and Rees L. H. (1983) The neuroendocrinology of opioid peptides. Br. bled. Bull. 39, 83-88. 16. Herbert J. (1981) Hormones and the sexual strategies of primates. In Perspectives in Primate Biology (eds Ashton E. H. and Holmes R. L.), Symp. zool. Sot. Lond. 46, 337-359. Academic Press, London. 17. Herbert J., Keverne E. B. and Yodyingyuad U. (1985) Modulation by social status of the relationship between cerebrospinal fluid and serum cortisol levels in male talapoin monkeys. Neuroendocrinology (in press). 18. Jackson S., Kiser S., Corder R. and Lowry P. J. (1985) Pro-opiocortin peptides in rat and cerebrospinal fluid. Regulat. Pept. 11, 159-171. 19. Jolly A. (1972) The Evolution of Primate Behavior. Macmillan, New York. 20. Keveme E. B., Leonard R. A., Scruton D. M. and Young S. K. (1978) Visual monitoring in social groups of talapoin monkeys (Miopithecus talapoin). Anim. Behav. 26, 933-944. 21. Lord J. A. H., Waterfield A. A., Hughes J. and Kosterlitz H. W. (1977) Endogeneous opioid peptides: multiple agonists and receptors. Nature 267, 495499. 22. McLaughlin L., Lowry P. J., Ratter S. J., Besser G. M. and Rees L. H. (1981) Characterisation of the proopiocortin family of peptides in human cerebrospinal fluid. Neuroendocrinology 32, 209-212. 23. Mains R. E. and Eipper B. A. (1981) Differences in the post-translational processing of fi-endorphin in rat anterior and intermediate pituitary. J. biol. Chem. 256, 56835688. 24. Martensz N. D. (1985) Changes in the processing of b-endorphin in the hypothalamus and pituitary gland of female rats during sexual maturation. Neuroscience 16, 625-640. 25. Martensz N. D. and Herbert J. (1982) Drug-induced hyperprolactinaemia and discharges of luteinizing hormone evoked by oestrogen in ovariectomized rhesus monkeys. J. Endocr. 94, 111-122. 26. Martin R. D. (1981) Field studies of primate behaviour. In Perspectives in Primate Biology (eds Ashton E. H. and Holmes R. L.), Symp. zool. Sot. Lond. 46, 287-336. Academic Press, London. 27. Meller R. E., Keveme E. B. and Herbert J. (1980) Behavioural and endocrine effects of naltrexone in male talapoin monkeys. Pharmac. Biochem. Behav. 13, 663-672. 28. Meyerson B. J. (1981) Comparison of the effects of f3-endorphin and morphine on exploratory and socio-sexual behaviour in the male rat. Eur. J. Pharmac. 69, 453-463. 29. Mohs R., Davis B. M., Rosenberg G. S., Davis K. L. and Kreiger D. T. (1982) Naloxone does not affect pain sensitivity, mood or cognition in patients with high levels of fl-endorphin in plasma. Life Sci. 30, 1827-1833. 30. Nakao K.. Nakai Y.. Oki S.. Matsubara S.. Konishi T.. Nishitani H. and Imura I. (1980) _ I Immunoreactive 8-endornhin in human’cerebrospinal fluid. J. clin. End&r. Metab. ‘50, 23Ck233. 31. Nakao K., Oki S., Tanaka I., Horii K., Nakai Y., Furui T., Fukushima M., Kuwayama A., Kageyama N. and Imura I. (1980) Immunoreactive /I-endorphin and adrenocorticotropin in human cerebrospinal fluid. J. clin. Invest. 66, 1383-1390. 32. Pezalla P. D., Lis M., Seidah N. G. and Chretien M. (1978) Lipotropin, melanotropin and endorphin: in uivo catabolism and entry into cerebrospinal fluid. J. Canad. Sci. Neurol. 5, 183-188. 33. Przuntek H., Stasch J.-P., Graf M., Ptlughaupt K. W., Gropp N. and Witteler M. (1981) A simplified radioimmunological method for the determination of human fi-endorphin in cerebrospinal fluid. J. Neurof. 224, 203-210. 34. Roberts A. C., Martensz N. D., Hastings M. H. and Herbert J. (1985) Changes in photoperiod alter the daily rhythm of pineal melatonin content, hypothalamic /I-endorphin content and the LH response to naloxone in the male Syrian hamster. Endocrinology 117, 141-148. 35. Roberts A. C., Hastings M. H., Martensz N. D. and Herbert J. (1985) Naloxone induced secretion of LH in the male Syrian hamster: modulation by photoperiod and gonadal steroids. J. Endow. 106, 243-248. 36. Sirinathsinghji D. J. S. (1984) Modulation of lordosis behavior of female rats by naloxone, /I-endorphin and its antiserum in the mesencephalic central gray: possible mediation via G&Ii. Neuroendoerinofogy 38, 222-230. 37. van Vugt D. A. and Meites J. (1980) Influence of endogenous opiates on anterior pituitary function. Fed. Proc. 3, 2533-2538. 38 Wardlaw S. L. and Frantz A. G. (1983) Brain B-endorphin during pregnancy, parturition and the postpartum period. Endocrinology 113, 16641668. 39. Wilkes M. M., Watkins W. B., Stewart R. D. and Yen S. S. C. (1980) Localisation and quantitation of @ndorphin in human brain and pituitary. Neuroendocrinology 30, 113-121. 40. Wuster M., Schulz R. and Herz A. (1983) A subclasai&ation of multiple opiate. receptors by means of selective tolerance development. J. Recep. Res. 3, 199-214. 41. Zakarian S. and Smyth D. G. (1982) p-Endorphin is processed differently in specific regions of rat pituitary and brain. Nature 2%, 250-252. (Accepted 27 January 1986)