Isolation of a Specific p-Opiate Receptor Peptide, Morphiceptin, from an Enzymatic Digest of Milk Proteins*

Specific radioimmunoassays have been developed for the measurement of naturally occurring morphiceptin and 8-casomorphin. These peptides and related exorphins were isolated from an enzymatic digest of caseins by chromatographic techniques including gel filtration, hydrophobic column and multiple-step high pres- sure liquid chromatography. Three exorphins were purified and characterized in their radioimmunologi- ca1, biological, and chemical properties. They were identified as morphiceptin, @-casomorphin, and &pro- lyl-@-casomorphin. Since morphiceptin is a highly specific p-agonist and can be derived from a milk protein, it is possible that morphiceptin is an exogenous opioid ligand specific for p-receptors in the brain and gas-trointestinal tract. reverse-phase chromatography (HPLC) em- for further purification. The in ml of 20 mM ammonium acetate (pH 4.2) and to a Magnum 9 2 column (Whatman, Partisil PXL 10/50 ODs) and eluted with a linear 2 acetonitrile gradient (10 to 50%) in 20 mM ammonium acetate buffer c (pH 4.2) for 2 h at 2 ml/min. Active fractions were pooled, lyophilized separately, and reapplied to the same column. These materials were further purified by reverse-phase HPLC on an analytical column (Partisil PXL 10/25 ODS) with a 0 to 50% acetonitrile gradient (30 min) and on ion-exchange HPLC columns (Partisill0 SAX/25). The z identity of each peptide was confirmed using synthetic standards. These final homogenous materials were subjected to gel filtration (Bio-Gel P-2, 1 X 22-cm column), receptor binding assay, and mass spectrometry. Opiate receptor binding assays were performed with rat (male Sprague-Dawley) brain membrane preparations using '=I-FK 33824 (0.05 nM, 2 Cilpmol) and [~-Ala*,~-Leu~]enkephalin (0.05 nM, 2 ci/ pmol) as labeled p-agonist and &agonists, respectively (9, 10). The purified peptides were hydrolyzed in 6 M HCl at 130 "C for 16 h and subjected to amino acid analysis. The results are expressed as ratios relative to phenylalanine content. The nearest integral values are also shown. 10 pA. Argon was employed as the collision gas in the middle quadrupole of this instru- ment. Reported experiments were run on two separate peptide preparations. In the first run, a copper target was employed, while in the second run stainless steel was used. No differences were observed with the different target materials.

Specific radioimmunoassays have been developed for the measurement of naturally occurring morphiceptin and 8-casomorphin. These peptides and related exorphins were isolated from an enzymatic digest of caseins by chromatographic techniques including gel filtration, hydrophobic column and multiple-step high pressure liquid chromatography. Three exorphins were purified and characterized in their radioimmunologi-ca1, biological, and chemical properties. They were identified as morphiceptin, @-casomorphin, and &prolyl-@-casomorphin. Since morphiceptin is a highly specific p-agonist and can be derived from a milk protein, it is possible that morphiceptin is an exogenous opioid ligand specific for p-receptors in the brain and gastrointestinal tract.
We have recently reported that a synthetic tetrapeptide amide fragment (Tyr-Pro-Phe-Pro-NH2) of a milk protein, @casein, is a potent and specific agonist for p-opiate receptors (1). This peptide was named morphiceptin. It has potent analgesic and cataleptic activities in rats when administered intracerebroventricularly (2, 3) and produces bradycardia by intravenous administration (4). These effects are blocked by naloxone, suggesting p-opiate receptor-mediated effects. Brant1 et al. (5) first isolated @-casomorphin, a heptapeptide (Tyr-Pro-Phe-Pro-Gly-Pro-Ile), from a peptone digest of /3casein and showed its opiate-like activity in inhibiting the electrically stimulated muscle contraction of isolated guinea pig ileum (5, 6). Morphiceptin is an amino-terminal tetrapeptide fragment of @-casomorphin, but the former is about 50 to 100 times more active than the latter in receptor binding assays (1). Although the physiological significance of morphiceptin is not known, the peptide may be important if proven to exist naturally. The present studies establish a possible natural existence of morphiceptin. It can be isolated from an enzymatic digest of casein. The isolated material was characterized as morphiceptin from its radioimmunoactivity, receptor binding specificity, and by mass spectrometry. In addition, @-casomorphin and 8-prolyl-@-casomorphin were isolated from the same digest.

MATERIALS AND METHODS
Synthetic morphiceptin, 8-casomorphin, and other related peptides were synthesized by the solid-phase method and were available from previous studies (1, 2). Enzymatic casein hydrolysate (lot 40F-0560) was purchased from Sigma. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adoertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Antisera against morphiceptin and P-casomorphin were raised in male New Zealand White rabbits. The peptide was conjugated to bovine serum albumin (Pentex, 5 X crystallized) with 1% glutaraldehyde in 0.25 M NaPO. buffer (pH 7.4) for 15 min at room temperature. The conjugate was separated from unreacted peptide by gel filtration on a Sephadex G-25 column. The void volume fractions containing conjugate (200 pg of peptide) were emulsified with an equal volume of complete adjuvant and injected into rabbits subcutaneously on both sides of the body and intramuscularly in both hind paws. The rabbits were boosted every 3 weeks with the conjugate (about 50 pg of peptide) emulsified with incomplete adjuvant.
A radioimmunoassay (RIA') using protein A containing Staphylococcus as the solid-phase support (7) was employed. Specific anti-Tris. HCl buffer (pH 7.7) containing 0.5 mM EDTA and 0.5% Triton serum (0.5 ml) was incubated with washed Staphylococcus in a 50 mM X-100 for 3 h to absorb most IgG. The IgG-absorbed Staphylococcus was then washed twice with the same buffer. The final pellet was suspended in five times the original serum volume and stored at 4 "C.
RIA was carried out in 0.25 ml of a 50 mM Tris.HC1 buffer containing 0.5 mM EDTA and 0.5% Triton X-100 for 2 h at room temperature or for 18 h at 4 "C with about 10,000 cpm of '251-labeled peptide (2 Ci/pmol) and 50 p1 of 100-fold diluted Staphylococcus'IgG complex (final antiserum dilution, 500-fold). After incubation, 2 ml of the same buffer was added to each tube, which was centrifuged for 20 min at 3,500 rpm in a Sorvall RC-5 centrifuge (rotor GSP-25). The supernatant was aspirated. The tubes containing Staphylococcus were counted in a y-counter at 75% efficiency. Nonspecific binding (less than 5%) was determined until an excess of unlabeled peptide (10 p~) .
8-Casomorphin-like and morphiceptin-like activities were determined from standard curves using synthetic peptides as standards.
The isolation procedure is depicted in Scheme 1. Ten grams of casein digest (Sigma) was dissolved in 20 ml of 0.1 M acetic acid by brief boiling. This solution was cooled to 4 "C and the precipitate that formed was removed by centrifugation. The clear supernatant was applied to a Bio-Gel P-2 column (2.8 X 53 cm) pre-equilibrated and eluted with 0.1 M acetic acid. Five-ml fractions were collected. Aliquots were assayed by RIA for morphiceptin-like and 8-casomorphinlike activities. The major 8-casomorphin-like and morphiceptin-like activities were separately pooled and lyophilized. The dried materials were dissolved in HzO and adsorbed batchwise on Amberlite XAD-2.
Most of the casomorphin and morphiceptin absorbed to the beads. These were washed extensively with 1 M acetic acid and the peptides were then eluted with methanol containing 1 M acetic acid. The eluates were dried under nitrogen and lyophilized. Semipreparative The abbreviations used are: RIA, radioimmunoassay; HPLC, high pressure liquid chromatography. reverse-phase high pressure liquid chromatography (HPLC) was employed for further purification. The residues were dissolved in 0.5 ml of 20 mM ammonium acetate (pH 4.2) and applied to a Magnum 9 2 column (Whatman, Partisil PXL 10/50 ODs) and eluted with a linear 2 acetonitrile gradient (10 to 50%) in 20 mM ammonium acetate buffer c (pH 4.2) for 2 h at 2 ml/min. Active fractions were pooled, lyophilized Z separately, and reapplied to the same column. These materials were 8 further purified by reverse-phase HPLC on an analytical column (Partisil PXL 10/25 ODS) with a 0 to 50% acetonitrile gradient (30 8 min) and on ion-exchange HPLC columns (Partisill0 SAX/25). The z identity of each peptide was confirmed using synthetic standards. These final homogenous materials were subjected to gel filtration (Bio-Gel P-2, 1 X 22-cm column), receptor binding assay, and mass spectrometry.
The purified peptides were hydrolyzed in 6 M HCl at 130 "C for 16 h and subjected to amino acid analysis. The results are expressed as 2 80 ratios relative to phenylalanine content. The nearest integral values are also shown.

60
Mass spectra were recorded on a Finnigan triple-stage quadrupole mass spectrometer (11) equipped with a Saddle-Field neutral beam 8 40 gun from Ion Tech Ltd., Teddington, England. The gun was operated z at 6 keV with xenon as the discharge gas. Discharge current was approximately 10 mA and the beam current was 10 pA. Argon was 20 employed as the collision gas in the middle quadrupole of this instrument. Reported experiments were run on two separate peptide preparations. In the first run, a copper target was employed, while in the 9 8 7 6 5 4 second run stainless steel was used. No differences were observed with the different target materials.

RESULTS
RIA of Morphiceptin and @-Casomorphin-Specific antibodies to morphiceptin and @-casomorphin were raised in rabbits. The specificity of the anti-morphiceptin antibody is shown in Fig. 1. Morphiceptin binds to this antibody with a high affinity. A median inhibitory concentration (IC, value) of 3 nM in displacing 1251-morphiceptin from the antibody was obtained under standard conditions. This same antibody shows less than 0.1% cross-reactivity with other related peptides such as @-casomorphin, @-casomorphin-5 (Tyr-Pro-Phe-Pro-Gly), Des-Tyr-@-casomorphin, and the tetrapeptide acid (Tyr-Pro-Phe-Pro) (Fig. lA). Met-and Leu-enkephalins show virtually no cross-reactivity (data not shown). 8-Casomorphin and 8-casomorphin-5 show 0.01 to 0.02% crossreactivity. This may account for the activity detected in those fractions on HPLC which do not correspond to morphiceptin.
Isolation of Morphiceptin-The procedure for the isolation of morphiceptin is summarized in Scheme 1. An acetic acid extract of casein hydrolysate was chromatographed on Bio-Gel P-2 (Fig. 2). The peak of @-casomorphin activity precedes the morphiceptin activity. Fractions 33-55 were pooled for the purification of 8-casomorphin and morphiceptin. The pooled fractions (33-55) were absorbed on Amberlite XAD-2 and eluted with methanol. This methanol extract was dried under nitrogen, lyophilized, and subjected to preparative HPLC. Four major peaks of morphiceptin-like material were detected. The largest peak, which corresponds to morphiceptin, was found in fractions 23-28 (Peak 11). The minor peaks, fractions 16-18 (Peak I), [33][34][35][36], and 39-42 (Peak IV) were identified later (Fig. 3). Fractions from Peak I1 were rechromatographed on Magnum 9 and then on an analytical reverse-phase and ion-exchange column (Fig. 4). Both mor-  Bio-Gel P-2 Amino acid analysis Mass spectrometry Receptor binding assay SCHEME 1 phiceptin radioimmunoactivity and preceptor binding activity were determined for each fraction from the analytical reverse-phase column (Fig. 5). Activities coincided exactly at the same fractions and eluted with the same retention time FIG. 2. Gel filtration of morphiceptin and 8-casomorphin immunoreactivity from a casein digest. The digest (35 g) was applied to Bio-Gel P-2 (2.8 X 53 cm) and eluted with acetic acid. Five-ml fractions were collected. An aliquot of each fraction was assayed for morphiceptin (0) and P-casomorphin (0) immunoactivities using the standard RIA described under "Materials and Methods." Absorbance of the eluate was monitored at 280 nm (A).

FIG. 3.
Separation of morphiceptin-like and casomorphin-like material from pooled fractions 33-55 of Fig. 2 by semipreparative reverse-phase HPLC. A Magnum 9 column (Whatman, Partisil PXL 10/50 ODS) was used with a 0 to 50% gradient of acetonitrile in 20 mM ammonium acetate buffer (pH 4.2) for 2 h at 2 ml/min to elute the peptides. Four-ml fractions were collected and an aliquot of the diluted sample was assayed for morphiceptin (0) and 0-casomorphin (0) activity by RIA. Absorbance of the eluate was monitored at 280 nm (A).
It should be pointed out that because of the small quantity of morphiceptin in our preparation we have been extremely careful to avoid any contamination by synthetic morphiceptin at any step. In all chromatography and HPLC, columns were extensively washed and samples injected only after we had not detected any morphiceptin-like activity in a prior control run. Thus, it is unlikely that the purified material is contam-FRACTION NO. The total morphiceptin-like activity in 10 g of the casein digest was about 40 nmol. The final purified material, containing about 5 nmol(12% yield) of morphiceptin equivalent, was subjected to further characterization by mass spectrometry.
Isolation of p-Casomorphin-A similar procedure was employed to obtain P-casomorphin, and the morphiceptin-like P-casomorphin-like materials from the other peaks seen on semipreparative HPLC (Fig. 3). Two peaks of immunoactivity eluted from Magnum 9 with retention times corresponding to the two minor morphiceptin-like activities (Peaks 111 and IV of Fig. 3) after morphiceptin. Both were purified to homogeneity with Magnum 9 and an analytical reverse-phase HPLC column and characterized. The total p-casomorphin-like activity in 10 g of casein digest is about 18 pmol, 400 times more than that of the morphiceptin in Peak 11. Peak I11 material, when eluted from 0.02 Absorbance of the eluate was monitored at 280 nm (---).  an analytical reverse-phase column, has a retention time corresponding to synthetic p-casomorphin, while Peak IV material is eluted behind p-casomorphin. The specific activity of the @-casomorphin of Peak I11 is consistent with that of (3casomorphin; Peak IV has 0.3% cross-activity based on absorbance at 280 nm.

FIG. 5. Elution profiles of RIA and radioreceptor activity ( R R A )
Peak I material was also purified to homogeneity by a similar procedure. This material showed very poor receptor binding activity and was not further characterized.
Comparison of the Materials from Peaks 11, 111, and IV-Elution profiles for morphiceptin and P-casomorphin were compared on analytical reverse-phase HPLC. Retention times of 12, 14.5, and 18.5 min (Fig. 7) were observed for Peaks 11, 111, and IV, respectively. Peak I1 corresponds to synthetic morphiceptin, Peak I11 to P-casomorphin (Fig. 7 A ) . The mo-  lecular sizes of Peaks 11, 111, and IV were compared by gel filtration (Bio-Gel P-2, 1 X 22 cm) and are in the order IV > I11 > 11, consistent with molecular weights.
Peaks I11 and IV were found to be 50 to 100 times less potent than morphiceptin in displacing "'I-FK 33824 from preceptors (Fig. 6). The potency of Peak I11 is similar to that of synthetic 8-casomorphin. Peak IV seems slightly more potent than Peak 111.
Amino Acid Compositions-The purified materials from Peaks I11 and IV were subjected to amino acid analysis. The results are shown in Table I. Peak I11 contains Tyr:Phe: Pro:Gly:Ile (1:1:3:1:1) and Peak IV Tyr:Phe:Pro:Gly:Ile (1:1:4:1:1). Peak I11 was 8-casomorphin and Peak IV 8-prolyl-/3-casomorphin. These assignments were confirmed by mass spectrometry. at m/z 522 whose collision-activated dissociation spectrum compares well with that in Fig. 9. The differences were due to concentration and matrix effects and because the mass window in the first quadrupole was wider for the isolate than for the standard.

Mass
Peak I11 Table I1 and is consistent with the structure of /3-casomorphin.
column was used. The flow rate was 2 ml/min, a 10 to 50% acetonitrile gradient was employed for 30 min, and 1-ml fractions were collected.   (Table 111).

DISCUSSION
Large numbers of biologically active neural or endocrine peptides terminate with an a-amide group at the COOH terminus, for example, corticotropin-releasing factor, gastrin, secretin, vasoactive intestinal peptide, cholecystokinin, growth hormone-releasing factor, substance P, bombesin, neurokinins, melanocyte stimulating hormones, and thyrotropin-releasing factors (12)(13)(14)(15)(16)(17). For most of these peptides, the COOH-terminal amide is required for optimal biological activity (12, 13). Two COOH-terminally amidated opioid peptides (metorphamide and amidorphin) have also recently been isolated from adrenal medulla and brain (18)(19)(20). The pancreas also produces a family of a-amidated peptides, the pancreatic polypeptides (i.e. NPY and PPY) (21). An enzyme  with the ability to convert peptides to a-amidated peptides has recently been identified from pituitary glands (22,23) and tumor cells (24). This enzyme is selective for peptides in which a neutral amino acid is followed by glycine (22-24). This enzyme is present in the secretory granules of pituitary cells and can be released upon stimulation (24). Since aamidated peptides are widely distributed in the gastrointestinal tract, including the pancreas, a similar enzyme may be present in the pancreatic enzyme mixture which was used to digest casein. Moreover, residues 60-64 of bovine ,&casein (Tyr-Pro-Phe-Pro-Gly) match the specificity for this enzyme, and it is conceivable that morphiceptin is produced from 0casein or its fragments (i.e. P-casomorphins) by a similar aamide converting enzyme. With the large quantity of 8casomorphin and 8-prolyl-0-casomorphin isolated from the digest, it is most likely they originate from /%casein in consistence with studies reported by Brantl et al. (3, 5, 6). The present study demonstrates the possible natural existence of morphiceptin in casein hydrolysate and confirms the existence of P-casomorphin in a peptone digest of casein (5, 6). Based on immunoactivity, receptor specificity, HPLC and gel filtration profiles, amino acid composition, as well as mass spectrometry, the three purified peptides are identified to be morphiceptin, /3-casomorphin, and 8-prolyl-~-casomorphin. Morphiceptin has the highest p-receptor binding activity (1). 8-Casomorphin is present in the largest quantity in the casein digest. According to radioimmunoassays, the total morphiceptin-like activity is 43 nmol/lO g of digest, whereas the total P-casomorphin-like activity is 18 pmol/lO g of digest, a quantity about 400-fold that of morphiceptin. However, morphiceptin has a higher p-receptor affinity than 0-casomorphin.

Isolation of Morphiceptin from Casein Digest
Because of the high selectivity for p-receptors in the brain (l), morphiceptin may play a unique function, in contrast to enkephalins which have been shown to prefer &receptors (9, 25). Morphiceptin and its analogs have been shown to produce analgesia, catalepsy (2, 3), and physical dependence in vivo (26), and to inhibit intestinal smooth muscle contraction (1). It is tempting to speculate that these exorphins may play a role in the infant-mother bonding relationship that is essential for the survival of infants. Recently, other exogenous opioids from food sources have been described (27)(28)(29). The demonstration that dietary sources of exorphins may produce effects in vivo (28,30) also suggests that exorphins such as morphiceptin and P-casomorphin could have physiological roles.