Molecular and Cellular Endocrinology, 48 (1986) 161-166 Elsevier Scientific Publishers Ireland, Ltd.

161

MCE 01558

Electrophysiological test of an amphiphilic /3-structure in LHRH action

Jenn-Tser Pan, Lee-Ming Kow *, Debra A. Kendall, Emil Thomas Kaiser and Donald W. Pfaff

T?w Rockefeller University, 1230 York Avenue, New York, NY I0021 (U.S.A.)

(Received 1 May 1986; accepted 28 July 1986)

Key words: arcuate nucleus; ~~vent~cul~/suprac~asmatic preoptic nucleus; brain slice; single-unit activity; norepinephrine; serotonin.

Summary

Micropipettes were used to record electrical activity from single neurons in hypothalamic tissue slices, in the hypothalamic arcuate nucleus (ARC), and in the periventricular and suprachiasmatic preoptic nuclei (POA). Responses were measured following in vitro application of luteinizing hormone releasing hormone (LHRH), and two analogues: LHRH models 1 and 3. Model 1 (pyroGlu-His-Trp-Ser-Phe-Thr-Ile-Lys-Ile- ThrNH,) had amino acid substitutions in residues 5-10 designed to form an amphiphilic P-strand structure. Model 3 (pyroGlu-His-To-Ser-Phe-Gly-Ile-Lys-Pro-SerNH~) was also designed to possess ~phip~lic characteristics, but also more closely to resemble the native peptide. Electrical recording results showed that LHRH was able both to excite or inhibit different hypothalamic neurons, and that it was more effective in the preoptic area than in the arcuate nucleus. Responses to LHRH model 3 were strongly correlated with responses to LHRH, in their occurrence and their direction, for both of the brain regions studied. Moreover, LHRH models 1 and 3 also retained the neuromodulatory effects of LHRH on cellular responses to norepinephrine and serotonin. Thus, LHRH analogues, designed to possess an amphiphilic p-structure, preserved some of the properties of LHRH when tested electrophysiologicahy in the central nervous system.

Since its isolation and synthesis in the early 197Os, luteinizing hormone releasing hormone (LHRH) has been studied extensively in brain and pituitary. LHRH analogues have proven valuable for clinical as well as physiological investigations (Bex and Corbin, 1984). Since LHRH can facili- tate mating behavior (Moss and McCann, 1973; Pfaff, 1973) even in hypophysectomized animals, actions of the decapeptide on other neurons are of

* To whom all correspondence should be addressed.

interest. LHRH analogues have been compared in behavioral assays (Kastin et al., 1980; Dudley et al., 1982). For analyzing the responses of individ- ual neurons, however, electrical recording is more valuable; here we have compared responses by neurons in the arcuate nucleus of the hypothala- mus (ARC) and the periventricular and supra- chiasmatic preoptic nuclei (POA) to LHRH and specific analogues in vitro, using rat brain tissue slices.

Although numerous analogues of LHRH have been synthesized and assayed for activity, correla- tions between primary sequence specificity, sec- ondary structure, and biological function remain

0303-7207/86/$03.50 0 1986 Elsevier Scientific Publishers Ireland, Ltd.

Fig. 1. The amino acid sequences of LHRH and models 1 and

3. The diagram shows the orientation of hydrophobic residues

(enclosed in boxes) and hydropic residues in an ~p~p~ic

~-~nfo~ation.

unclear (Crighton et al., 19Sl). The amino acid sequence of LHRH suggests that this hormone could interact with an amphiphilic target site (e.g. a membrane surface) by forming an amphiphilic #&structure. In this conformation, alternating side chains would project above and below the plane of the peptide backbone such that one face of the molecule would be comprised of the hydrophobic residues and the other face the hydrop~lic res- idues. This possibility is supported by circular dichroism analysis which indicates that in increas- ingly nonpolar solvents, LHRH undergoes a tran- sition from a random coil to a more ordered &conformation (Mabrey and Klotz, 1976).

Accordingly, the main purpose of the present study was to investigate the role of an amphiphilic &conformation in LHRH activity. For this, we have synthesized analogues which are optimized for formation of this structural unit yet which retain minimal sequence homology with the natu- ral peptide (Fig. 1). These analogues are compared with LHRH for electrophysiological activity. A similar approach for analogue design has previ- ously elucidated the role of ampbiphilic a-helices in other peptide hormones (Kaiser and Kezdy, 1983).

Methods

Adult female Sprague-Dawley rats ovariecto- mized for at least one week were used. Half of the animals were implanted S.C. with 5 mm long estradiol-filled Silastic tube at the same time

ovariectomy was done. All estrogen-treated rats showed es~ogen-dependent lordosis behavior, and enlarged uteri while none of the nontreated rats did so.

Rats were sacrificed by decapitation and brains were taken out of the skull within 2 min for trimming and slicing. Thin (400 pm) coronal sec- tions containing the preoptic (POA) or ARC hy- pothalamic region were prepared with a Vibra- tome (Lancer, series lOOO), and stored for 1 h before recording began in artificial cerebrospinal fluid (ACSF) gassed with 95% 0, and 5% CO,. The composition of the ACSF is as follows (in mM): NaCl, 125.1; KCl, 3.8; KH,PO,, 1.2; MgSO,, 1.3; CaCl,, 2.4; NaHCO,, 26; dextrose, 10. To begin recording, a slice was laid on a nylon net immersed in a chamber which was continu- ously perfused with gassed ACSF at the rate of 1.5-2 ml/mm. The temperature of the chamber was maintained at 33°C. Extracellular single-unit activity was recorded conventionally with an ACSF-filled micropipette (5-10 MOhm), which was guided visually into the desired area under a dissecting microscope, using a ~crom~ipulator

When a neuron showed stable baseline activity, it was subjected to testing with LHRH and LHRH analogues. A neuron was often also characterized by its responses to neurotransmitters such as norepinephrine (NE), serotonin (5-HT), acetyl- choline (ACh), histamine (Hist); and peptides such as thyrotropin-releasing hormone (TRH) and cholecystokinin-octapeptide sulphated (CCK-8). All the test agents were dissoIved in normal saline, and 50 ~1 of each was injected directly into the chamber above the inlet at a time. The concentra- tions used for each agent are listed in the figure legends. Consecutive app~cations were made at least 5 min apart, after the firing rate returned to baseline level or reached a new steady state. For more detailed description, please consult Pan et al. (1986).

LHRH models 1 and 3 were synthesized using solid-phase techniques on a benzhydrylamine res- in for incorporation of the carboxy-terminal amide (Pietta and Marshall, 1970). The peptides were cleaved from the resin and the protecting groups removed using hydrogen fluoride in dimethyl sulfide (Tam et al., 1983). Purification of each peptide was accomplished using HPLC with an

acetonitrile gradient in 0.1% aqueous TFA. Purity was determined by the presence of a single, sym- metrical peak upon elution from a reverse-phase Cl8 column and then verified by amino acid analysis.

The synthesis and physical characterization of the peptide (Val-Glu-Val-(TFA)Om),-Val, de- noted here as p-9, were described previously (Os- terman and Kaiser, 1985). This peptide possesses an amphiphilic P-strand structure but has no primary sequence homology with LHRH.

ReSUltS

Neurons in the preoptic area tended to be more responsive than those from the arcuate nucleus of the hypothalamus to LHRH (&i-square test, P < 0.001) as well as LHRH model 3 (P -c 0.02) (Table 1). In each of the two cell groups, percentage distributions across the types of responses were similar between LHRH and its analogue model 3. Indeed, when tabulated neuron by neuron, for all cells tested with both the native peptide and LHRH model 3, when LHRH evoked a particular type of response, analogue model 3 showed a significant tendency to yield the same response both in the arcuate nucleus (P < 0.01) and in the preoptic area (P < 0.01) (Table 2).

In addition to transmitter-like effects on neuro- nal activity (Fig. 2), LHRH can also modulate the effects of classical transmitters such as norepi- nephrine and serotonin (Pan, Kow and Pfaff, manuscript submitted). The neuromodulatory ac- tions of LHRH can also be shown by the ana- logues used here. For example, in Fig. 3B, both

TABLE 1

163

TABLE 2

RESPONSES OF ARCUATE HYPOTHALAMIC (ARC)

AND PREOPTIC (POA) NEURONS TO LHRH AND LHRH

MODEL 3 APPLICATION, FOR CELLS THAT WERE

TESTED WITH BOTH PEPTIDES

LHRH3 LHRH Total

+ 0

ARC neurm~~

+ 7 0 3 10 - 0 4 0 4

0 1 0 52 53

Total 8 4 55 61

POA neurons

+ 14 0 5 19 - 0 3 0 3

0 I 0 41 48

Total 21 3 46 70

+ , excitation; - , inhibition; 0, no change.

LHRH and analogue model 3 had a positive neu- romodulatory effect on the response to norepi- nephrine. Similarity between the native peptide and its analogue was found even when the neuro- modulatory effect was negative (Fig. 3D). The most striking example is in Fig. 3C, in which LHRH and analogue models 1 and 3 all had a similar neuromodulatory effect on the action of serotonin: they increased the latency and de- creased the magnitude of the serotonin effect on this neuron. As a control the peptide p-9, which possesses an idealized amphiphilic B-structure was used, and it had no effect, indicating that the neuromodulatory effects of the other models were specifically due to their similarity to LHRH.

RESPONSIVENESS OF ARCUATE HYPOTHALAMIC (ARC) AND PREOPTIC (POA) NEURONS TO LHRH AND LHRH

MODEL 3

Brain Peptides Number of Percentage (W) of neurons showing areas neurons

tested + f 0

ARC LHRH 72 18 6 0 76

LHRH3 15 13 I 0 80

POA LHRH 250 45 7 2 46

LHRH3 66 35 6 0 59

+ , excitation; - , inhibition; f , biphasic response; 0, no response.

164

B

C D SNT LWtn LMRH3

3 ACh NE

2

:, t

Fig. 2. Examples of recordings from five preoptic area neurons, during and after applications of LHRH, LHRH model 3 and other agents. All traces display neuronal firing rate as a function of time. Arrows indicate the time points at which various agents were applied. The volume of each application was 50 ~1, and the calculated peak concentration of each agent in the bath was as follows: LHRH, LHRH model 3, TRH, 250 nM; norepinephrine, 12.5 pM; 5-HT, ACh, histamine, 25 CM. Neuron A responded to later applications of both LHRH and LHRH model 3 with larger amplitude and shorter latency responses, than to earlier applications. In neuron B, both LHRH and LHRH model 3 inhibited firing, to about the same degree. In neuron C, they both yielded small increases in firing rate, even though the neuron could be inhibited, as indicated by the response to ACh. In neurons D and E, LHRH and LHRH 3 did not give responses identical to each other. Responses to other agents are also shown, as controls. For exampIe, in neuron D, the excitatory response to histamine indicates that the lack of a response to LHRH model 3 was not because the neuron had lost its excitability.

A 3 2

(:

Discussion

With electrophysiological recording, as we have done, on large numbers of individual central nervous system cells, there was every opportunity for LHRH analogues to behave differently from the native peptide. Thus, it is interesting that the analogues devised to optimize an amphiphilic /3- structure proved to retain many transmitter-like and neuromodulatory actions of LHRH, and when an individual neuron was tested with both they tended to give similar types of response.

In turn, similarities between these analogues and the native peptide LHRH in the central nervous system do not necessarily imply that we should expect similarities in other target tissues. LHRH analogues 1 and 3 did not possess signifi- cant LH-releasing activity tested in vivo, nor did they bind to LHRH receptors on pituitary cell membranes (unpublished data). Contrasts of pitu- itary functions of LHRH analogues with behav- ioral effects in the central nervous system have already led to the conclusion that there are differ- ing receptor requirements for LHRH, between pituitary and brain. Kastin et al. (1980) and Sakuma and Pfaff (1983) found that neither the presence nor the direction of effect of an LHRH analogue in releasing LH predicted what would occur in behavioral tests. Similarly, Moss and Dudley (1984) showed that a fragment of the LHRH molecule inactive at the pituitary level was still capable of inducing behavioral receptivity in estrogen-primed female rats, when it was given in the cerebral ventricles. Our results suggest that recognition of LHRH by components of the central nervous system and the pituitary involves different degrees of sequence specificity. Since LHRH is only 10 residues long, it is quite possible

165

that primary sequence specificity and secondary structural features are superimposed. Optimiza- tion of the p-conformation in LHRH analogues, however, involves a concurrent loss of amino acid functional groups which appear specifically re- quired for pituitary action but not central nervous system activity.

Our results demonstrate that analogues which have little sequence homology with LHRH (i.e. 40% in model 1 and 60% in model 3) do possess LHRH-like neuronal activities. This may reflect a particular requirement for the first ‘four amino- terminal residues in the natural sequence (and which have been retained in the analogues) and may involve an amphiphilic /&conformation. As- says with the p-9 peptide indicate this structural feature alone, however, is insufficient for activity. These findings suggest that analogues, designed to vary the contribution of sequence homology and structural elements, provide a means for delineat- ing the mechanisms by which LHRH interacts with different target sites.

Acknowledgements

We thank Dr. D.G. Osterman for providing us with a sample of the 8-9 peptide and Mrs. Adelaide Acquaviva for typing the manuscript. This re- search was supported by NIH grant HD 05751 and Institutional Biomedical Research Grant No. 5-26954 (to D.W.P.), NIH postdoctoral fellowship HD 06570 (to D.A.K.), and USPHS Program Pro- ject grant HL-18577 (to E.T.K.).

References

Bex, F.J. and Corbin, A. (1984) In: Frontiers in Neuroendo-

crinology, Vol. 8, Eds.: L. Martini and W.F. Ganong

(Raven Press, New York) pp. 85-151.

Fig. 3. Examples of neuromodulatory effects of LHRH and its analogues, as well as transmitter-like responses in preoptic neurons.

Volumes and concentrations of agents used are the same as in Fig. 2. In addition, LHRH model 1, the control &pleated sheet

peptide, 8-9 and CCK-8 were used in concentrations.of 250 nM. Neuron A responded to LHRH model 1, but not to LHRH itself,

model 3, or p-9. Neuromodulation was shown in neuron B: LHRH slightly potentiated the action of norepinephrine on this neuron,

while LHRH model 3 had a pronounced effect on the response to norepinephrine. In neuron C, LHRH, LHRH model 1 and LHRH

model 3 all strikingly reduced the electrical response to 5-HT (serotonin), while the control, p-9, did not. Reduced responses to

serotonin following neuromodulation consisted both of a decreased amplitude of response and an increased latency. All neuromod-

ulatory effects recovered. Record is continuous from Cl to C2 (the horizontal bar at the end of Cl shows the overlap with the

beginning of C2). In neuron D, both LHRH and LHRH model 3 reduced the response to norepinephrine without suppressing

baseline electrical activity and without a nonspecific suppression of excitability (note response to CCK-8).

166

Crighton, D.B., Haynes, N.B., Foxcroft, G.R. and Lamming,

G.E. (Us.) (1981) Control of Ovulation (Butterworth, Bos-

ton).

Dudley, C.A., Vale, W., Rivier, J. and Moss, R.L. (1982)

Peptides 2, 393-396.

Kaiser, E.T. and Khdy, F.J. (1984) Science 223, 249-255.

Kastin, A.K., Coy, D.H., SchaIIy, A.V. and Zadina, J.E. (1980)

Pharmacol. Biochem. Behav. 13,913-914.

Mabrey, S. and KIotz, I.M. (1976) Biochemistry 15, 234-242.

Moss, R.L. and Dudley, C.A. (1984) In: Handbook of Psycho-

pharmacology, Vol. 18, Eds.: L.L. Iversen, S.D. Iversen and

S.H. Snyder (Plenum Press, New York) pp. 397-454.

Moss, R.L. and McCann, S.M. (1973) Science 181, 177-179.

Osterman, D.G. and Kaiser, E.T. (1985) J. Cell. B&hem. 29,

57-72.

Pan, J.-T., Kow, L.-M. and Pfaff, D.W. (1986) Neuroendo-

crinology 43, 189-196.

Pfaff, D.W. (1973) Science 182,1148-1149.

Pietta, P.G. and Marshall, G.R. (1970) J. Chem. Sot. D1970,

650-651.

Tam, J.P., Heath, W.F. and Merrifield, R.B. (1983) J. Am.

Chem. Sot. 105,6442-6455.