Photoinactivation of the μ opioid receptor using a novel synthetic morphiceptin analog

European Journal of Pharmacology, 139 (1987) 273-279

273

Elsevier EJP 00832

Photoinactivation of the/ opioid receptor using a novel synthetic mol hiceptin analog W.F. Herblin *, J.C. K a u e r a n d S.W. T a m Medical Products Department, E.I. du Pont de Nemours and Company, Wilmington, DE 19898, U.S.A.

Received 23 March 1987, accepted28 April 1987

The benzophenone chromophore has been incorporated into a synthetic amino acid (p-benzoyl-L-phenylalanine; L-Bpa) to produce a chemically stable photoaffinity probe. L-Bpa was found to retain the photochemical reactivity of benzophenone. To test the utility of this synthetic amino acid as a photo-reactive probe for receptors, a tetrapeptide analog of morphiceptin was made as a model peptide in which the C-terminal prolinamide was replaced by L-Bpa amide. The affinity of the ~t opioid receptor for this peptide is comparable to that for the parent compound, morphiceptin. Irradiation of the peptide-receptor complex reduced the subsequent binding of [3H]naloxone and virtually eliminated that of [3H]Tyr-D-Ala-Gly-NMe-Phe-Gly-ol (DAGO). Binding studies with [3H]naloxone indicated that both the affinity and the capacity were reduced. Competition studies with [3H]D-Ala2-D-LeuS-enkephalin (DADLE) and naloxone indicated selective inactivation of a ~t type opioid receptor. Opioid receptors; /~ Receptors; Photoaffinity labeling; p-Benzoyl-phenylalanine; (Peptide)

1. Introduction Photoaffinity labeling of receptors is an important tool in biochemical pharmacology, and it has been used in various forms to label many receptors (Fedan et al., 1984). It offers the advantage of lower non-specific reaction as compared to chemical affinity labeling and it offers greater specificity than chemical cross-linking. Azides are the most frequently used photo-labile group and have been used successfully in many receptor systems, including selective photolabeling of different opioid receptors (Hazum et al., 1979; Garbay-Jaureguiberry et al., 1984). However, the high reactivity of the azides limits their use to the final steps of synthesis of ligands. For the isolation of neuropeptide receptors and * To whom all correspondence should be addressed: E.I. du Pont and Company,ExperimentalStation, Bldg.400, Wilmington, Delaware 19898, U.S.A.

the study of peptide-protein interactions, a chemically stable amino acid photoaffinity probe which could be incorporated into peptide ligands by both solid phase and solution peptide syntheses is desired (Parker and Hodges, 1985). Additional desirable properties include (1) activation by low energy light (300-350 nm), (2) rapid reaction of the activated intermediate, (3) stability in an aqueous environment, and (4) probability of rapid decay to ground state in the absence of photochemical reaction (Cavalla and Neff, 1985). We have synthesized p-benzoyl-phenylalanine (Bpa), which contains the benzophenone chromophore and this amino acid seems to meet the above requirements (Kauer et al., 1986). On irradiation, the benzophenone portion of the Bpa molecule would be expected to form a free radical which could couple to adjoining peptide structures. We have therefore prepared an analog of morphiceptin in which the C-terminal prolinamide has been replaced by L-Bpa amide and tested this

0014-2999/87/$03.50 © 1987 Elsevier SciencePublishers B.V. (BiomedicalDivision)

274 peptide for opioid receptor binding as well as for photo-induced inactivation. The results reported here indicate that L-Bpamorphiceptin amide (4-BMA) has selective affinity for the /~ opioid receptor and can selectively inactivate that receptor on irradiation.

2. Materials and methods

2.1. Synthesis of L-Bpa4-morphiceptin amide (4BMA) 2.1.1. BOC-p-benzoyl-L-phenylalanine amide A solution of 3.69 g (10 mmol) of Boc-p-benzoyl-L-phenylalanine (L-Bpa) (Kauer et al., 1986) and 1.01 g of N-methylmorpholine in dry tetrahydrofuran was cooled to - 2 0 ° C and 1.36 g of isobutyl chloroformate was added, followed in 5 min by a solution of 1 ml of concentrated ammonium hydroxide in 1 ml of acetone. The mixture was stirred at 0 °C for 30 min, then solvent was stripped and the residue was shaken with 150 ml of water and 150 ml of ethyl acetate. The organic layer was washed with 5% aqueous citric acid; then with aqueous sodium bicarbonate and dried (MgSO4). Solvent was stripped and the residue (3.434 g) was twice crystallized from ethyl acetate to give 1.8 g of white crystalline BOCamide as a hemi-hydrate, mp 148-149°C. Anal. (C21H24N204 1/2 H20), calculated C, 66.82%; H, 6.69%; N, 7.47%. Found: C, 66.80%; H, 6.71%; N, 7.14%.

2.1.2. p-Benzoyl-L-phenylalanine amide acetic acid salt A 1.0 g portion of the BOC-amide was dissolved in 10 ml of 98% formic acid. After 2 h at 25 ° C, the solution was lyophilized. The residue was successively lyophilized from water and from acetic acid. This final residue was dissolved in 25 ml of methylene chloride and filtered to remove a small amount of p-benzoyl-L-phenylalanine. The solution was used as described in the next section.

2.1.3. BOC-L-tyrosyl(t-butylether)-L-prolyl-L-phenylalanyl-p-benzoyl-L-phenylalanine amide A suspension of 8.6 g (5 mmol) of oxime resin

(DeGrado and Kaiser, 1982), 2.652 g of BOC-Lphenylalanine and 5.0 ml of 2.0 M diisopropylcarbodiimide (DIC) in 96 ml of CH2C12 was shaken gently for 3 days. The resulting resin was filtered off and washed successively with 5 × 100 ml of 2 : 1 CH2C12/CEHsOH , 3 X 100 ml CH2C12, 3 x 100 ml CH3OH and dried. Unreacted sites were blocked by shaking for 15 min with 6.05 ml of C6HsCOC1 and 9.05 ml of diisopropylethylamine (DIEA) in 100 ml of CHEC12. The solid was filtered off and vacuum dried to give 10.1 g of BOC-L-Phe-oxime resin. A 4.69 g portion (1.37 mmol) of the resin was coupled successively with 3.6 mmol of the symmetric anhydrides of BOC-Lproline (3 h) and BOC-L-tyrosine (t-butyl ether) (17 h) using a Beckman 990B synthesizer and the program of Nakagawa and Kaiser (1983). The symmetric anhydrides were prepared by treating 7.2 mmol of the corresponding BOC-amino acid •(Bachem, Switzerland) in 38 ml of CHEC12 and 10 ml of DMF with 3.6 ml of I M DIC in CHEC12 at 0 o C. Note that there is no neutralization step in this program; after the addition of 3.6 mmol of the preformed BOC symmetric anhydride, 4.6 ml of 10% DIEA in CH2C12 was added. The resulting BOC-Tyr(tBu)-Pro-Phe-resin was washed with ethanol and after vacuum drying weighed 5.17 g. (Amino acid analysis found: Tyr, 0.301; Pro, 0.272; Phe, 0.292 mmol/g.) A 3.40 g portion of this resin was added to the p-benzoyl-L-phenylalanine amide acetic acid salt solution in methylene chloride (above) and 58/xl of acetic acid was added. The slurry was shaken gently for 40 h at 25 o C and filtered. The solid was washed with 2 × 15 ml CH2CI2, then with 4 x 25 ml CH3OH. Combined filtrates were stripped, the residue (1.423 g) in ethyl acetate was washed with 10% citric acid, 5% sodium bicarbonate, dried (MgSO4) and stripped. This residue (1.287 g) was triturated for 3 days with ether. The resulting solid (894 mg) was chromatographed in 75-120 mg portions on 30 g of Merck Kieselgel H using 3% ethanol in chloroform; 8 ml fractions were collected. Pooled fractions 23-26 yielded 480 mg of BOC-Tyr(tBu)-Pro-Phe-Bpa-NH 2. Rechromatography of fractions 27-32 produced an additional 104 mg, Rf 0.60 (Silica Gel H, chloroform/methanol 9 : 1). A 33 mg portion was crystallized from

275

1 ml ethyl acetate and 1 ml hexane to give 27 mg of white powder, mp 115-121. NMR 360 mHz (10% CD3OD in CDC13) 8 7.3-7.8 (m, 9H, Bpa aromatics), 7.1-7.3 (m, 5H, Phe aromatics), 6.89, 7.03 (dd, 4H Tyr aromatics, J = 8 Hz), 4.72 (m, H, Tyr et), 4.48 (m, H, Phe et), 4.35 (m, H, Bpa a), 3.35 (m, 2H, Pro 8), 2.9-3.2 (m, 4H, Tyr fl + Phe fl), 2.75 (m, 2H, Bpa fl), 1.75-2.0 (m, 4H, Pro fl +-/), 1.36 and 1.48 (d, 9H, Tyr OtBu), 1.31 (s, 9H, BOC CH3). Anal. (C48H57NsOs). Calculated N, 8.42%, found N, 8.42% (combined chromatography fractions 17-19 produced 75 mg of the deletion dipeptide BOC-Tyr(t-Bu)-Bpa-NH 2, NMR similar, but lacked Pro and Phe peaks.

2.1.4. L- Tyrosyl-L-prolyl-L-phenylalanyl-p-benzoylL-phenylalanine amide hydrochloride (4-BMA) A solution of 106 mg of BOC-Tyr(tBu)-ProPhe-Bpa-NH 2 in 3 ml of anisole was treated with 0.7 ml of 4 N HC1 in dioxane. The clear solution gradually deposited 104 mg of crystalline monohydrate which was separated by decantation, washed with ether, and vacuum dried. Amino acid analysis; found: proline, 1.038, tyrosine 0.881, phenylalanine 0.962. Anal. (C39H41NsO6HC1 H/O), calculated: C, 64.14%; H, 5.94%; N, 9.58%; found C, 63.75%; H, 5.69%; N, 9.58%. uv ~,m~(ethanol, C = 0.100 g/100 ml)hm~, 259 nm c 25000, 332 nm c = 160 (water, C = 0.100 g/100 ml) 265 nm ~16400, 335 nm (sh).

2.2. Binding studies The affinity of 4-BMA for various opioid receptor types was determined by the methods of Tam (1985) using guinea-pig brain membranes. Binding was performed in 50 mM Tris buffer with bacitracin (50 /~g/ml), pH 7.4, with or without 100 mM NaC1 under the following conditions: 0.5 nM [3H]naloxone (# binding); 1 nM [3H]DADLE in the presence of 4 nM sufentanil (8 binding); 1 nM [3H](-)-ethylketocyclazocine in the presence of 500 nM DADLE and 20 nM sufentanil (x binding). To examine the photoaffinity reaction, striatal tissue from male Sprague-Dawley rats was ho-

mogenized in 30 volumes of 0.32 M sucrose and centrifuged at 1000 × g for 15 min. The supernatant was then centrifuged at 40000 × g for 30 min to produce a crude synaptosomal fraction. The pellet was resuspended in the same volume of Tris-HC1 buffer (50 mM, pH 7.4) and incubated for 1 h at room temperature in the presence or absence of 2.8/tM 4-BMA. During the last 15 min of the incubation, half of each membrane preparation was irradiated in a Rayonet reactor containing 16 light tubes with a peak emission wavelength of 350 nm. All four membrane preparations were washed twice by centrifugation and resuspension in Tris buffer before use in binding assays. Binding was measured by incubating membranes (2.5 mg fresh tissue weight/ml) with 0.5-8 nM [3H]naloxone (42.7 Ci/mmol, New England Nuclear) 1-4 nM [3H]DAGO (40.1 Ci/mmol, New England Nuclear), or 0.5-4 nM [3H]DADLE (46.9 Ci/mmol, New England Nuclear) for 45 rain at room temperature, followed by rapid filtration through G F / B glass fiber filters under reduced pressure. The filters were washed with two 5 ml aliquots of cold buffer, dried and counted for tritium with an efficiency of 43%. For [3H]DAGO and [3H]naloxone binding, parallel samples containing 10 #M unlabelled naloxone were used to determine non-specific binding.

2.3. Data analysis Most of the data were analyzed using the computer program MLAB (Knott, 1979). [3H]Naloxone binding data were fit to a simple equation for a single binding site, since it was only necessary to determine whether or not the binding has been modified. Data for the binding of [3H]DADLE and its inhibition by naloxone were fit to an equation containing terms for the binding of two ligands to two independent sites and a term to describe non-specific binding.

XIJ il) + =2(L+K=+ 2XIJ i2)+Lx S

276 L = free ligand ( D A D L E )

SYNTHESIS OF L--BPA AMIDE

I = inhibitor (naloxone)

o

Kd] , Kd2 = dissociation constants for D A D L E

H2N-CH-COOH

HzN--CH-C--NHz

Kil , Ki2 = inhibition constants for naloxone NS = non-specific binding

I) 2) 5) 41 5)

C=O [~

3. Results

Me3COCOzN=C~ [sobOCOCI • NH3 HCOOH CH3COOH

C=O [~

L--BPA

L--SPA--NH z

(r)

(2)

3.1. Chemistry SOLID PHASE SYNTHESIS OF MORPHICEPTIN ANALOG (3)

Since it was desired to prepare several morphiceptin analogs with various C-terminal photoreacfive groups, the recently reported oxime resin technique (DeGrado and Kaiser, 1982) was employed. This resin when substituted with a protected peptide sequence is the equivalent of an active ester, and displacement of the peptide with an amine, amino ester or an amino acid amide leads to C-terminal substitution of the protected peptide with the corresponding amide, amino ester, or aminocarboxamide derivative. In this work, we employed the amide of the recently reported photoreactive amino acid p-benzoyl-L-phenylalanine (Bpa) (Kauer et al., 1986). The BOC derivative of Bpa (1) was converted via the mixed carbonic anhydride to the amide (2) (fig. 1). The protected tripeptide sequence BOCTyr(tBu)-Pro-Phe was synthesized on the oxime resin and treated with 2 in the presence of acetic acid (catalyst) to produce the protected tetrapeptide BOC-Tyr(tBu)-Pro-Phe-Bpa-NH2 and a small amount of the deletion dipeptide BOCTyr(tBu)-Bpa-NH 2. The protected tetrapeptide was purified by column chromatography and deprotected with 4 N HC1 in dioxane to give crystalline 4-BMA (3) as a pure hydrochloride monohydrate salt.

3.2. Biology The affinity of 4-BMA for opioid receptors was determined as described in the Materials and methods section (table 1). 4-BMA showed highest affinity for the /~ receptor (0.27 /~M) which is similar to that of morphiceptin (0.25 vM). The affinity for 8 or x receptors was considerably lower. The addition of 100 mM NaC1 to the

HON----~

,) (HBOC-P.,)zO

Z) 3) (BOC-Pro)20 4) H÷ 5) (BOC-Tyr{IBu|)20

BOC-TYr-Pro-Phe--ON=~ tBu /L--Bpa--NH2 (2) purified by chromatography, crystallization

BOC-TYr-Pro-Phe-Bpo--NH z tBu

J.cl

,o,

,o

,o,

o,

HzN- CH-C--N---t--C-NH--CH--C--NH-CH--C--N H2 t / \ I I CHz ~ CHz CHz

T

~ OH

~ T

C--'O

©

crystaline

hydrochloride monohydrote

Tyr-- Pro--Phe--Bpa--NH z (3)

Fig. 1. Outline of the steps in the synthesis of 4-BMA.

assay caused a slight reduction in the affinity for/~ and x receptors but the affinity for 8 receptors was unchanged. Preliminary results indicated that the degree of photo-inactivation was related to the concentraTABLE 1 Affinity of Tyr-Pro-Phe-BpaNH2 (4-BMA) for opioid receptors. Binding was performed in Tris buffer in the presence or absence of 100 m M NaC1. Values represent mean of triplicates + S.E.M.

(/~M)

Receptor

IC50

types

No NaC1

100 m M NaC1

/~ /~ K

0.268 + 0.082 3.67 +0.79 8.20 +0.57

0.590 -I-0.043 3.57 +0.5 17.3 +2.37

277

tion of the peptide as well as the length of irradiation. Standard conditions were chosen to minimize irradiation to prevent receptor damage and to approach saturation by the peptide (10 × Kd). To investigate the proposed photo-induced reaction of 4-BMA with the opioid receptor, the specific binding of [3H]naloxone to control and photo-reacted membranes was determined (fig. 2). The derived parameters from the best-fit, single site models indicate a reduction in Bmax and an increase in K d after photo-reaction. Membranes which had been irradiated in the absence of 4-BMA or incubated with the peptide without irradiation showed no significant differences from control (untreated) membranes, indicating that the irradiation alone did not inactivate the opioid receptor and that the washing conditions were sufficient to remove any residual effects of the peptide. When [3H]DAGO was used in similar experiments, the specific binding in the photo-reacted tissues was reduced by 90% or more. [ 3H]D-Ala2-D-LeuS-enkephalin ( D A D L E ) binding was performed with control and photo-reacted membranes to identify the type of opioid

0.020-

co- O o =

r.r3

0.010

o/I

+ + +

2000-

Z 0

" ..~-

+

~.-..... ~: ...

1500 -

* ............ ~/ ....

~. "'~"-. &..

"+ ............. +

1000 -

.....~ h" t. -.. zx

o b,-1

9"""0

0

[-,

.....

"'o

• .....

+

".A. ..

0

o'--...~ ......

• "l'""'o"O. .

a

.%.... z~

~"""o.

..... e--.. I...,..o

-I

0

t

- ................

2

o -c~........... e

.... • ............. •

3

4

LOG NALOXONE(taM) Fig. 3. Inhibition of the binding of [3H]DADLE to control brain membranes by naloxone. Membranes were incubated with 0.5 (o), 1 (O), 2 (zx) and 4 nM ( + ) [3H]DADLE in the presence of different concentrations of unlabeled naloxone.

receptors being inactivated. Four concentrations of [3H]DADLE were incubated with various concentrations of naloxone and the data fit to an equation for two independent sites plus nonspecific binding. If one of the Bm~x values is set to zero, this equation reduces to a single site model. The 'best fit' model for the data obtained with control membranes (fig. 3) was a two site model (F = 13, P < 0.001). Both sites showed relatively

:/

/.

2500-

/

/

Z

TABLE 2 m o

o.oo -

0.000

-

/

~/~PIOTO-REACTED

I

I

I

0

1

2

LOG [3H]NALOXONE(]aM) Fig. 2. Binding of [3H]naloxone to control and 4-BMA-photoreacted brain membranes. Membranes were incubated with 0.5-8 nM [3H]naloxone. Binding was performed as described in Materials and methods.

Binding of [3H]DADLE to brain membranes. Naloxone was

used to compete with the binding of 0.5, 1, 2 and 4 nM [3H]DADLE, and the binding parameters of each ligand were determined by computer analyses to fit a 2-site model. Site 1, with higher affinity for naloxone, resembles a ~ receptor, and site 2, with high affinity for DADLE and low affinity for naloxone, resembles a 8 receptor.

Bmax fmol/mg tissue

K d (DADLE) K i (naloxone)

Site 1

Site 2

683 c.p.m. 6.1 2.8 nM 2.0 nM

515 c.p.m. 4.6 1.2 nM 53 nM

278

2000-

+

÷

~ ; ~L +

la.,

"'~,.

+

+

1500 -

+ +

Z 0 IzI

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

4-

A

1000 -

& /k .............. ~' ~'---.& ,, 2"...R

A '.~...

[...,

5oo ......

8

O

~"

0 0 .~......8.... 8

~

A

.......

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

O-

l

I

I

l

I

-

0

1

2

3

4

LOG N A L O X O N E

(nM)

Fig. 4. Inhibition of the binding of [ 3 H ] D A D L E to 4-BMAphoto-reacted brain membranes by naloxone. Membranes were incubated with 0.5 (C)), 1 (e), 2 (zx),and 4 n M (+) [3H]DAD L E in the presence of different concentrations of unlabeled naloxone.

high affinity for DADLE (/~ and 8), but only one of these sites showed high affinity for naloxone (~t) (table 2). When membranes which had been irradiated in the presence of the peptide were used for the binding of [3H]DADLE and its inhibition by naloxone, the data were best fit by a single site model with Bm~x = 502 c.p.m., K d (DADLE) = 1.7 nM, and K i (naloxone) = 12 nM (fig. 4). Although the data were more variable than was seen with control membranes, there was no suggestion of a second site. The high affinity for DADLE and low affinity for naloxone of this site resembles a 8 receptor, and the Bm~x is virtually identical to the site observed in the control membranes (table 2).

4. Discussion

Although the replacement of proline by L-Bpa amide at the C-terminal position of morphiceptin slightly decreased the affinity for the /~ opioid receptor, the resulting peptide still binds with appreciable potency and retains the specificity for

the/~ receptor shown by the parent morphiceptin. Irradiation of the peptide-receptor complex produced a reduction in the subsequent binding of [3H]naloxone which was characterized by a decrease in both the affinity and the capacity. Since naloxone is known to bind to multiple opioid sites, one interpretation is that a high-affinity site for naloxone has been inactivated. This is supported by the virtual elimination of [3H]DAGO binding, a more selective ligand. The results with [3H]DADLE binding also confirm this interpretation. In control membranes, [3H]DADLE binds to two naloxone-sensitive sites which show relative affinities resembling # and 8 receptors. After photo-reaction of membranes with the Bpa-containing peptide 4-BMA, only one site is detected and its high-affinity for DADLE and low affinity for naloxone suggest that it is a 8 site. Therefore, the /L site observed on control membranes has been selectively inactivated. While 4-BMA is of interest as the representative of a new class of photo-reactive probes, its low affinity for the opioid receptors presents several problems. The low affinity of 4-BMA requires the use of high concentrations which increases the possibility of non-specific reaction and makes experiments to protect against the photoreaction difficult because of the possibility of direct reaction with the protecting ligand. It is our intent to emphasize the approach rather than the specific compound. We have found that the benzophenone chromophore, incorporated into an amino acid (L-Bpa), retained its photochemical reactivity and shown that a model peptide, a morphiceptin analog incorporating L-Bpa in the C-terminal position, retained selective opioid receptor binding affinity. Irradiation of the peptide when bound to the opioid receptor produced inactivation of the receptor which was apparently specific for the # type opioid receptor. The use of L-Bpa has very broad applications, since it can be easily incorporated into peptides during their syntheses. The suitability of L-Bpa-containing peptides as photoaffinity probes for peptide receptor isolation remains to be determined by future studies with higher affinity L-Bpa-containing peptides. The development of an amino acid photoaffinity probe

279

may well have general utility in the study of other peptide receptors and peptidases. References Cavalla, D. and N.H. Neff, 1985, Chemical mechanisms for photoaffinity labelling of receptors, Biochem. Pharmacol. 34, 2821. DeGrado, W.F. and E.T. Kaiser, 1982, Solid-phase synthesis of protected peptides on a polymer-bound oxime: preparation of segments comprising the sequence of a cytotoxic 26peptide analogue, J. Org. Chem. 47, 3258. Fedan, J.S., G.K. Hogaboom and J.P. D'Donnell, 1984, Photoaffinity labels as pharmacological tools, Biochem. Pharmacol. 33, 1167. Garbay-Jaureguiberry, C., A. Robichon, C. Dauge, P. Rossignol and B.P. Roques, 1984, Highly selective photoaffinity labeling of tt and 8 opioid receptors, Proc. Natl. Acad. Sci. U.S.A. 81, 7718.

Hazum, E., K.-J. Chang, S. Wilkinson and P. Cuatrecasas, 1979, Fluorescent and photoaffirtity enkephalin derivatives: preparation and interaction with opiate receptors, Biochem. Biophys. Res. Commun. 88, 841. Kauer, J.C., S. Erickson-Viitanen, H.R. Wolfe, Jr. and W.F. DeGrado, 1986, p-Benzoyl-L-phenylalanine, a new photoreactive amino acid: photolabeUing of calmodulin with a synthetic calmodulin-binding peptide, J. Biol. Chem. 261, 10695. Knott, G.D., 1979, MLAB-A mathematical modeling tool, Comput. Progr. Biomed. 10, 271. Nakagawa, S.H. and E.T. Kaiser, 1983, Synthesis of protected peptide segments and their assembly on a polymer-bound omixe: applications to the synthesis of a peptide model for plasma apolipoprotein A-l, J. Org. Chem. 48, 678. Parker, J.M.R. and R.S. Hodges, 1985, Photoaffinity probes provide a general method to prepare synthetic peptide-conjugates, J. Prot. Chem. 5/6, 465. Tam, S.W., 1985, (+)-[3H]SKF 10,047, (+)-[3H]ethylketocyclazocine, It, K, 8, and phencyclidine binding sites in guinea pig brain membranes, European J. Pharmacol. 109, 33.