Brain Research, 586 (1992) 303-310 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006.8993/92/$05.00

303

BRES 17936

Ontogenesis and binding properties of high-affinity neurotensin receptors in human brain N i c o l e Zsiirger a, JoSlle C h a b r y a, A n t o i n e C o q u e r e l b and J e a n - P i e r r e V i n c e n t a a institut de Pharmacologie Mol~culaire et Cellulaire du CNRS, Valbonne (France) and t, Laboratoire de Radioanaly~c CHR de Rouen, Rouen (France) .

(Accepted 3 March 1992)

Key words: Neurotensim Receptor; Ontogenesis; Human brain

The ontogenesis of neurotensin binding sites was studied in human brain of subjects deceased from Sudden Infant Death Syndrome. Monoiodo-Tyr3neurotensin specifically recognized 2 distinct classes of binding sites in human brain homogenate. The high affinity sites were already present at birth and increased to a maximal level of 240 fmol/mg protein 1 month after birth. Thereafter, the density of these sites decreased to reach a value of 8 fmol/mg protein in 15-month-old brain, a value similar to that found in adult brain. The dissociation constant of the high-affinity sites (about 0.3 nM) did not vary from birth to adulthood. The high-affinity binding sites were sensitive to GTP which decreased their affinity for neurotensin by a factor of 3, indicating that these sites are functional receptors coupled to GTP-binding proteins. By contrast, the low-affinity sites were insensitive to GTP and could be partly blocked by the antihistaminic drug levocabastine. These sites were absent in human brain during the first post-natal year and could be detected only in brain homogenate of 15-month-old infants. The transient increase in high-affinity neurotensin binding sites after birth suggests that neurotensin could act as a regulatory peptide during brain development.

INTRODUCTION Neurotensin is a tridecapeptide isolated both from mammalian brain .~,44 and gastrointestinal tract m~..,0..~.,. In the central nervous system (CNS), numerous anatomical ~7,41 and pharmacological results suggest that neurotensin modulates the activity of dopaminergic neurons (for review, see ref. 22). For example, neurotensin binds to receptors located on dopaminergie neurons both in rat 23,4oand in human brain ~,~,.~8. The use of iodinated neurotensin iigands with high specific radioactivity led to the detection in murine brain of 2 components of neurotensin binding sites with different binding parameters and pharmacological properties 2s.30.45. The high-affinity (K,j ranged from 0.05 to 0.5 nM) and low-capacity (Bmax ranged from 10 to 60 fmol/mg) binding sites mediate the intracellular response to neurotensin 2-4; the physiological function of the low-affinity (K o = 2-8 riM) and high-capacity

(100-150 fmol/mg) binding sites remains unknown 40, A high-affinity binding site has been purified to homogeneity from 8.day-old mouse brain 2,~.This site seems to be different from the high.affinity neurotensin receptor that has been cloned from a eDNA library of adult rat 42, Ontogenesis of both high- and low-affinity sites were studied in routine brain. Total neurotensin binding sites increased from embryonic day 14 to post-natal day 8 then decreased to a 4-fold lower value in adult brain 3.~. Data concerning neurotensin binding sites in the human brain are less abundant because human tissues are not easily available and also because adult brain contains very low amounts of neurotensin binding sites 37,3~. The present study describes the ontogenesis of neurotensin binding sites in the human brain during the first 15 months of postnatal life. The relative abun-

Correspondence: J.-P. Vincent, Institut de Pharmacologic Mol~culaire et Cellulaire du CNRS, 660 Route des Lucioles - Sophia Antipolis, 06560 Valbonne, France. Fax: (33) (93) 95 77 08.

304 dance of high-affinity binding sites in 2-month-old human brains made it possible to characterize their structural and pharmacological properties. MATERIALS AND METHODS

Source of brain tissue Post-natal ontogenesis studies of neurotensin binding sites in human brain were carried out with brains from 2 contol infants and 7 subjects who died of Sudden Infant Death Syndrome (SIDS). This pathology is not associated with neurocortical disease but could be related to immaturity of medulla centers that control breathing (for review, see refs. 18. 27) The SIDS diagnosis was established on the basis of negative necropsic examination and clinical investigations consisting of bacterial and viral analysis and interview of parents• in the course of these interviews, we pointed out the lack of antecedent of convulsion. dsspnea, vomiting and the absence of fever or acute disease during the days preceding the death. The sex. age. cause of death and associated symptoms of each case are summarized in Table 1. Ages at death ranged from I to 410 days. Although the 15-month-old case 9 was diagnosed its a typical SIDS. This case is rarc since only 2c~" of SIDS are above I year of age is. Autopsy delay did not exceed 48 h and had no significant effect on the binding properties of brain tissue ~'"~' Preparation of tissue honlogenates All steps were performed at 4°C. One human hemisphere without brainstem (1(~)-300 g wet tissue) was homogenized using a Polytron homogenizer for 30 s in 10 vols, of ice-cold Tris-HEPES buffer 5 mM, pl=l 7,5, containing a mixture of peptidase inhibitors (5 mM EDTA, 0,1 mM phenymethylsulfonyl fluoride, I mM iodoacctamide aitd I ~M pupstatin), llomottenatcs were centrifuged (30 rain, IIHI,IHIOx g), the supernatant wits discarded and the pellet wits washed twice in the s:tme buffer. The final pellet wits resuspended (20 mg/ml) in Tris.ItEPES buffer 20 raM, pl=l 7,5, containing I()C~ glycerol (w/v) anti the same protease inhibitors, Protein concentra. tion was determined by the Warht, rg and Christian procedure 4r, using a Beckman spcctrophotometer, ltomogenates were stored at = I Sl)"(', Mouse and rabbit brain homottenat~s were prepared from Iotal brain without cerebellum accordi,g to the procedure described above, Bindit#~ c'Xl~t'rhtwttts Ilomogenate (0,1 mr; protein) was incubated with 0,1 nM it. lrlyrrneurotensm (20(H) Ci/mmol) for 2() rain at 25°C in 25{)/.tl of 5{) mM Tris-tlCI buffer. DH 7.5, containing ().2~ bovine serum albumin, 0•1 ~M N-henzyloxycarbonyl prolyl prolinal and 5 mM EDTA, Binding experiments were terminated by addition of 2 ml

ice-cold buffer. Radioactivity bound to homogenate was separated from free ligand by filtration under reduced pressure through cellulose acetate filters (Sartorius, 0.2 ~.m). Filters and tubes were washed twice with 2 ml of incubation buffer. Radioactivity retained on filters was counted with a 3' counter (Packard, counting efficiency 80%). Non-specific binding was measured in the presence of an excess of unlabeled neurotensin (1 ~tM) and was always lower than 5% of total binding. High pressure liquid chromatography analysis showed that at least 90% of the radiolabeled ligand used for this experiments remained intact. Association kinetic of [t251]Tyr3-neurotensin to neurotensin receptors was started by addition of homogenate to 0.1 nM of radiolabeled ligand. Bound radioactivity was measured at different times by filtration. Binding equilibrium was attained after 15 rain at 25°C. The dissociation kinetic was initiated by addition of I ~tM unlabeled neurotensin and was followed by measuring the radioactivity bound to homogenate at different times. Saturation experiments were performed by incubating membranes with increasing concentrations of [t2Sl]Tyr~-neuroiensin alone (from 25 to 400 pM) or isotopically diluted by unlabeled neurotensin (from 0.1 to 10 nM)). Binding parameters (dissociation constant K d and maximal binding capacity Bmax) for a single site or for"2 different types of sites were derived from computerized Scatchard analysis t3. Competition experiments were performed by incubating membranes with increasing concentrations of various synthetic and natural derivatives of neurotensin (from 10-i-. to 10 -¢' M). IC5o values were calculated from inhibition curves as peptide concentrations inhibiting 50% of [t's,,.-. "" qzyr.~-neurotensin binding. To investigate the eff,.~t of cations and nucleotides, the homogenate was washed twice with a modified binding buffer that did not contain EDTA, then it was incubated with increasing concentrations of various cations and nucleotides in the same buffer. All cations were used as their chloride salt.

RESULTS

Otttogetwsis of netlrotensin hireling sites Fig. IA shows that only high-affinity binding sites were detected from birth to post-natal month 15. High-affinity binding sites were present at birth (B,.,, x - 5 0 fmol/mg). Their density increased to reach a peak level of 240 fmol/mg I month after birth (Fig. 2C,D). Then the amount of high-affinity binding sites decreased and reached the adult value 15 months after birth (8 fmol/mg, see ref. 37). These variations of binding capacity occurred without any noticeable change of the affinity constant (Kit = 0.3 + 0.05 nM).

TABLE I

('hantctetistic,~, of the st.died st~bjec'ts Std~j('¢t

Sex

!

F

2 3 4 :'

M F M F

c~

F

7

M

"J

M M

Age at the time of death 1,5 h

4 days 15 days 32 days 32 days 60 days 90 days 145 days 15 months

C,~,~,wof death

A,v,~¢a~iatedsymptoms

Cardiorespiratory failure Heart failure SIDS SIPS SIPS SIDS SIDS SIDS SIDS

Premature (31 weeks); Sacrococcygeal teratoma Cardiac malformation Minor dysembryoplasy Bronchitis Benign cervical dysplasy Hypotrophy None Minor dysembryoplasy Ectopic fundus: Liver lymphatic distension

:i

305

AI

100

-

•

er-te

T.

I'~

~

• • Time(rain)

r

100

so

~ 75 Time(mini

2"/

e

o" so

' ,sL"". q

t

•,

T

[ ,-r"-, j

/'%

t

I

I

I

0

5

10

15

,, ~"

al [,.

21 ~

I;

1.6

_~

1.4

.~

1.41

.

o

.

,

J S'

2

' 60 0

10

20

30

Time (min)

OtOO I

0

0

I

0

5

0

I

0

I

10

~"

lS

'' 60

Age (days) Fig. !, Comparative ontogeny of the high- and low-affinity neumtensin binding sites in mammalian brain. A: human brain from non-SiPS ( I , r a ) or SIPS (e,o), B: mouse brain. C: rabbit brain. Maximal binding capacity of high-affinity (e) and low-affinity (o) binding sites were determined from Scatchard plots as described in Fig. 2, A: each point is the mean :i: S,E.M. of triplicate experiments carried out with a single brain homogenate. B and C: each point is the mean :i: S.E.M. of triplicate experiments carried out with three different brain homogenatcs.

?,S

\~

(T)emna,sa fmollm"

8 3,S

I °[m%~Z--';',, ;.s = e,s

0

10

3O

3O

Bmax:230fmol/mo-

m

so

3O

10 50 F InU)

lOO 200 e (fmol/mg)

0

Fig. 2. Binding of ['2Sl]Tyrrneurotensin to human brain, Brain homogenate (0.1 mg protein) was incubated at 25°C with i,~creasing concentrations of [tZ~l]Tyrrneurotensin in the presence (non-specific binding, - ) or in the absence (total binding, o) of I #M ncurotensin. The specific binding (e) was calculated as the difference between total and non-specific binding. Each point is the average of duplicate determinations from a representative experiment. Saturation curves (A, C) and scatchard plot analysis (B, D) were obtained with homogenates of 15-month-old (A, B) or I-month-old (C, D) human brains. B and F, bound and free concentrations of ligand.

Fig. 3. Association and dissociation kinetics for the binding of [l'~Si]Tyrrneurotensin to homogenates of I-month-old human brain. The association reaction (e) was followed by measuring the amount of [t~'~i]Tyr.rneurotensin (0.1 nM) specifically bound to brain homogenates (0,1 mg protein) after various incubation times at 2YC. Inset a: pseudo-first-order represention of association data. Dissociation kinetic (o) was initiated by I /zM unlabeled neuro~ensin. Inset b: first-order represention of dissociation data.

Low-affinity binding sites were absent during the first post-natal year and were first detected in 15. month-old human brain (Fig. IA). Both the affinity and the binding capacity of these sites were low and therefore difficult to characterize accurately. Estimated values were about 3 nM for affinity and 30 flnol/mg for binding capacity (Fig. 2A,B). Ontogenesis of neurotensin receptors in mouse (Fig. I B)and rabbit brain (Fig. IC)were studied fo,' comparison. In mouse brain, high-affinity binding sites are also present at birth, increased to a peak value on post-natal day 8 (250 fmol/mg protein) and decreased to reach the adult value on day 60 (50 fmoi/mg protein). Mouse low-affinity binding sites first appeared at post-natal day 10 ano\,orogressively increased to reach a plateau value at day t~0. In rabbit brain we could not detect any low-affinity bifiging sites. Binding capacity is maximal at birth and deci~ased by a factor of 3 from day 7 to adulthood.

Properties of the high aJJ,,my binding sites in two.monthold brains Kinetics of assocLqtion and dissociation. Binding of [t251]Tyr3-neurotensin to l-month-old brain homogenate was time-dependent and reversible. Specific binding was maximum after 3 rain at 25°C (Fig. 3). Non-specific binding remained constant and lower than 5% of total binding. The free [IzSI]Tyr3-neurotensin concentration varied less than 10% during the course of the association. Therefore the association kinetic was pseudo-first-order and a semilogarithmic plot of

3O6 sic cleavages of neurotensin are inactivating since corresponding partial sequences are less potent than neurotensin itself, 3.8% efficiency for neurotensin 9-13 and less than 0.1% for neurotensin 1-8 and neurotensin 1-10. Natural analogs of neurotensin-like xenopsin, neuromedin N or LANT6 are substituted in the 8-13 C-terminal sequence. Their potencies relative to neurotensin were found to be between 3% and 12%. These results show that the integrity of the 8-13 Cterminal sequence is required for optimal binding of neurotensin to its high-affinity sites in human brain. Effects of cations and nucleotides. Low concentration of Mg z+ (from 0.1 to 1 raM) resulted in a slight increase of neurotensin binding between 10 to 40%. Concentrations of Na +, Li +, Mg z+, and Ca '+ higher than 2.5 mM decreased binding of radiolabeled neurotensin with a common ICs0 value of 25 raM; sensitivity to K + is lower since a 150 mM concentration is necessary to obtain 50% inhibition of binding.

the data was linear (left inset in Fig. 3). The value of the observed rate constant of association w a s kap p = 0.4 rain -I. The expression of kapp is: k,mp=k,,(i'-~lTyr.~ neurotensin) + k a, where k.., and k a represent respectively the second-order rate constant of association and the first-order rate constant of dissociation of the [t,..Si]Tyr3.neurotensin_receptor complex. Dissociation was initiated by addition of 10 -¢' M neurotensin (Fig. 3). Because the presence of a large excess of unlabeled neurotensin prevented the reassociation of [l:'~l]Tyr.~neurotensin to its receptor, the dissociation process is first order with k d as rate constant. The kd value (0.3 rain- =) was calculated from the slope of the semilogarithmic representation of the dissociation kinetic (right inset in Fig. 3). From values of k,,pp and k d it is possible to calculate k,,= 1 nM - I ' m i n -I and the affinity constant Kd = k,,/k,, = 0.3 nM. Smctural bimling specificity. Specificity of neurotensin binding was studied in competition experiments involving truncated sequences, synthetic derivatives and natural analogs of neurotensin. Results are summarized in Table !i. Analogs that were shortened at the N-terminal end like neurotcnsin 2-13 and 8-13 were slightly more potent than ncurotcnsin. Acetylneurotensin 8-13 wits approximately its potent as neurotensin itself. On

GTP, GDP and the nonhydrolysable GTP analog GTP-yS decreased the binding of [IZSl]Tyr3-neurotensin to its human receptor in a concentration-dependent manner. The order of decreasing inhibitory potency was OTP-yS > GTP = GDP. Other nucleotidcs, like ATP, ADP, AMP, CTP and GMP had no inhibitory effect. Scatchard analysis of saturation experiments performed in the presence of 10 =s M GTP-TS showed that the nucleotide decreased the affinity by a factor of 3 without affecting the maximal binding capacity (Fig, 4A), These experiments required the presence of 2 mM Mg: * for a maximal sensitivity to GTP

the other hand. analogs substitued in C-terminal region were less potent than the parent peptide, The analog in which Tyr ** wits replaced by a Trp wits 15-fold less potent than neurotcnsin, Neurotensin I=10 an~ 1=8 exhibit very low affinities for the human neurotensin r~:ceptor with ICs. value higher than I() p.M, Peptida-

1'ABLE II

('onlpclriSOtl or

t/it.

hOIdoIg pOlOlt'y

Of II('tlrOl('tl,~'itl COldit,~ ,~lrllt'llll'(li COlCl/O,l~lW,~/

I('~, h~M~

( bmi~mml

Seqtwnc,,

N~:urotensin Neurotensin(8-13) Neurotensin(2-13) AccIylN'I'(8-13) [TrpZZJNT Xetmpsin Ncurotensin Iq-I.~) NeuromedinN

¢)E L Y E N K P II R k Y E N K P R acR DE L V E N K P R pE L

LANT, [DTrpZZ]NT Neurotrnsin(I-8) Neuroten.~in(I-IO)

[DLeu*t][Trp**INT

pE ¢)E DE pS

L k L L

Y Y Y Y

E E E E

N N N N

R R R It R R # k ] k N g P R R K P II K P I~ R K P R R

P P P P P P

P P P P P P

Y Y V Y W W V Y Y dW

! X ! X I X

X I

I I

k L k k k k k k L k

g ][ dL

0,2(0.2) 0.08 I1.1.t, 0,4

3 .~.2 5.2 6.8(I.6)

14 (3.4) 150 > I000 > IO(X) >l(g~(i

Potency I(XI(I(X)) 250 154 50 7 3,8 3.8 2.9(12.5)

1.4(5.9) 0.I