Isolation of a SiaIogogic Peptide from Bovine Hypothalamic Tissue and Its Characterization as Substance

SUMMARY A sialogogic peptide has been isolated from bovine hypothalami by gel filtration, ion exchange chromatography, and high voltage paper electrophoresis. The purity of this peptide was established by showing that the molar ratios of the constituent ammo acids remained constant after high voltage paper electrophoresis at pH 1.9, 3.5, and 6.5. Its molecular weight (estimated by gel filtration) and its amino acid composition show that the peptide is an undecapeptide composed of Lysl, Argl, Glx~, Proz, Gly,, Leur, MetI, and Phez. The NH2-terminal residue is arginine. Further biological testing showed that the sialogogic peptide isolated is indistinguishable from Substance P; i.e. it stimulates contraction of the guinea pig ileum and the rat duodenum and is a potent vasodepressor. These effects are not inhibited by prior treatment of the test preparation with atropine. This report is believed to present the first complete purification of Substance P.

A sialogogic peptide has been isolated from bovine hypothalami by gel filtration, ion exchange chromatography, and high voltage paper electrophoresis.
The purity of this peptide was established by showing that the molar ratios of the constituent ammo acids remained constant after high voltage paper electrophoresis at pH 1.9, 3.5, and 6.5. Its molecular weight (estimated by gel filtration) and its amino acid composition show that the peptide is an undecapeptide composed of Lysl, Argl, Glx~, Proz, Gly,, Leur, MetI, and Phez. The NH2-terminal residue is arginine. Further biological testing showed that the sialogogic peptide isolated is indistinguishable from Substance P; i.e. it stimulates contraction of the guinea pig ileum and the rat duodenum and is a potent vasodepressor.
These effects are not inhibited by prior treatment of the test preparation with atropine. This report is believed to present the first complete purification of Substance P.
In 1931 von Euler and Gaddum (1) showed that extracts of various equine tissues, especially brain and intestine, contained a substance that lowered arterial blood pressure when injected intravenously into anesthetized rabbits and stimulated the contraction of intestinal tissue in vitro. Neither of these effects was inhibited by preliminary treatment with atropine. They referred to their partially purified preparation as Powder P and this original designation evolved into its present name of Substance P. Subsequent workers have shown many other interesting biological effects of preparations of Substance P, in particular, effects on the central and peripheral nervous systems (2, 3). In 1967, Leeman and Hammerschlag (4) found that extracts of bovine and rat hypothalami contain a peptide that * This is Publication 732 from the Graduate Department of Biochemistry, Brandeis University, Waltham, Massachusetts 02154. This work was supported in part by National Institutes stimulates salivary secretion when injected into anesthetized rats. Shortly afterward, Lembeck and Starke (5) reported on the sialogogic activity of their preparations of Substance P and suggested the possible identity between Substance P and the sialogogic peptide.
This report presents a method for obtaining the sialogogic peptide in pure form, its amino acid composition, and further data comparing its biological activity and chemical properties to preparations of Substance P reported by other workers (2, S-10).

Analytical Procedures
Protein determinations were done by the method of Lowry et aZ. (11). Effluents from column chromatographic procedures were monitored by measuring the absorbance of the solution at 260, 280, or 300 rn+ Measurements of optical density of solutions of the sialogogic peptide were done with a Cary model 15 recording spectrophotometer with matched sets of quartz cuvettes for test and reference solutions. All amino acid analyses were done on a Beckman-Spinco automatic amino acid analyzer, model 12OC, according to the general procedures of Spackman, Stein, and Moore (12). The analyzer was adapted for high sensitivity by the addition of a 4-to 5.1-millivolt range in the recorder. Hydrolyses of the samples were performed for 24 hours at 110" with 0.2 ml of constant boiling 5.7 N HCI in vacuumsealed tubes. 1 -Dimethylamino -naphthalene -5 -sulfonyl (dansyl) chloride was used to determine the NHt-terminal residue according to the method of Gray (13). Electrophoresis was carried out at 80 volts per cm on a Michl type electrophoresis apparatus (14). The buffer systems were as follows: (1) pH 1.9: formic acid-acetic acid-water (20:80:900), (2) pH 3.5: pyridine-acetic acid-water (4:40 : 760)) (3) pH 6.5 : pyridineacetic acid-water (80:2.4:720), (4) pH 8.9: 1 Y0 NHdHC03 in water. Samples were applied to Whatman No. 1 chromatographic paper in 50 to 200 ~1 of 0.1 M acetic acid and 10u2 M mercaptoethanol. A mixture containing 10 to 20 mpmoles each of a mixture of standard amino acids was spotted on both sides of the samples. Staining was done by immersing the paper in a solution of ninhydrin-cadmium acetate reagent (15). To recover samples, horizontal strips of paper 2 to 3 cm wide were cut and eluted with 2 to 3 ml of 0.1 M acetic acid-lo-* M mercaptoethanol. Aliquots of the eluted samples were diluted with 0.9% NaCl before testing for sialogogic activity.

Biological Procedures
Testing for sialogogic activity was done according to the method of Leeman and Hammers&lag (4). A sialogogic dose was defined as that amount of material that stimulates the secretion of 50 =t 20 ~1 of saliva within 2 min following intravenous injection into rats weighing approximately 100 g and anesthetized with pentobarbital, 50 mg per kg. The contractility of freshly dissected sections of guinea pig ilia or rat duodena in response to added samples of peptide was measured in a 40-ml bath of Tyrode's solution (16) at 37", aerated with a mixture of air-CO2 (95 :5%).
The records were made by attaching the muscle to a lever and recording the deflections on a kymograph ( The frozen hypothalami were homogenized to a uniform consistency in a Gilford wood colloid mill in a solution (a volume was used which was 5 times in liters the frozen weight of the tissue in kilograms of The two filtrates were pooled and extracted five times fifth tube were diluted to 1 ml with 0.9% NaCl and tested for sialogogic activity. with petroleum ether (b.p. 3650.9").
Each time a volume of petroleum ether was used which was one-fifth of the volume of philized residue was rechromatographed on a smaller column of the combined filtrates.
The aqueous phase which contained all Sephadex G-25 maintained at room temperature (Fig. 2). The of the sialogogic material and some residual acetone was flash-active fractions were pooled and lyophilized. evaporated at room temperature and lyophilized. Extraction Cation Exchange Chromatography-The lyophilized residue and gel filtration of batches of hypothalami larger than 10 kg after the second gel filtration step was taken up in a sufficient were performed at the New England Enzyme Center, Tufts volume of 0.4 M pyridine-acetate buffer, pH 5.5, such that the University, School of Medicine, Boston, Massachusetts.
resulting solution had an absorbance at 300 mp of approximately Gel Filtration-The lyophilized hypothalamic extract was 10.0. This solution was applied to a column of sulfoethyl taken up in 0.1 M acetic acid and centrifuged at 10,000 x g for Sephadex C-25 previously equilibrated with 0.5 M pyridine-ace-20 min before applying the supernatant fluid to a column of tate, pH 5.5, maintained at room temperature. A stepwise Sephadex G-25 (fine) equilibrated with 0.1 M acetic acid at room elution with 170 ml of 0.50 M pyridine acetate at pH 5.5 followed temperature.
Material always was applied in a volume which by 1000 ml of 0.75 M pyridine-acetate, pH 5.5, was used (Fig. 3). was less than 5% of the bed volume of the column.
The elution Active fractions from the eluate from the column of sulfoethyl profile obtained after the extraction of 2 kg of bovine hypo-Sephadex were pooled and lyophilized. The residue was then thalami is shown in Fig. 1. For batches of 20 kg of hypothalami, dissolved in a volume of 0.005 M ammonia sufficient to raise the the lyophilized extract first was applied to a large column of pH of the solution to above 9.0 before applying it to a column of Sephadex G-25 (14 X 150 cm), maintained at 4". Fractions of carboxymethyl cellulose-11 (Whatman) that had been equili-250 ml were collected at a flow rate of 40 ml per min. The brated previously with 0.005 M ammonia. A stepwise elution column was screened for sialogogic activity by taking 0.2-to of 150 ml of 0.005 M ammonia followed by 160 ml of 0.04 M 1.0-ml aliquots of every third fraction for testing.

Chromatography
on Cm-cellulose of sialogogic material after sulfoethyl Sephadex chromatography.
Applied material was pooled from tubes 36 to 40 of Fig. 3. Column size: 1.4 X 15 cm. Fraction size: 8 ml. Flow rate: 1 ml per min. Aliquots of 20 to 50 ~1 from every tube were diluted to 1 ml with 0.9% NaCl and tested for sialogogic activity. raphy on Cm-cellulose,1 the effluent fractions with sialogogic activity were pooled and lyophilized, and the residue was applied to Whatman No. 1 paper in a 5-to lo-cm band. Electrophoresis was performed at pH 1.9 or 3.5 as the final step in purification.

Enzymatic Studies
The enzymes used were obtained from Worthington.
Trypsin, two times recrystallized, was treated with L-(l-tosylamidoSphenyl)ethyl chloromethyl ketone according to the method of Carpenter (18) to inactivate any trace of chymotrypsin. Samples of 10 sialogogic doses each of the sialogogic material obtained after chromatography on Cm-cellulose were incubated in 1 ml of buffer for 3 hours at 37" with either 10 pg of an enzyme or 10 pg of an enzyme that previously had been put in a boiling water bath for 10 min. The solution used for the hydrolysis with trypsin was 0.5% NHdHCOa; with pepsin, 0.01 N HCl; with chymotrypsin, 0.05 M Tris-HCl (pH 7.15); and with carboxy-1 The abbreviation used is: Cm, carboxymethyl.  A 3 X 1O-5 M solution of the pure peptide eluted from paper after electrophoresis at pH 3.5 in 0.1 N acetic acid-10m3 M mercaptoethanol gave essentially a zero absorbance in the 270 rnp to 300 rnp wave length region.
The reference solution was the eluant of an adjacent region preceding the sialogogic activity.

Estimation of Size
Samples of oxidized insulin A chain (mol. wt. 2284), arginine vasopressin (mol. wt. 1086), and sialogogic material at all stages of purity were run successively on the same column of Sephadex G-25 at room temperature (Fig. 6). From these data we conclude that the sialogogic peptide has a molecular weight greater than 1000 but less than 2000. This is in agreement with the molecular weight of 1340 calculated from the amino acid composition.
a Dose per analysis. b Starting material, bovine hypothalami obtair he( c Starting material, bovine hypothalami obtained from Swift. d These values do not include the methionine sulfoxide and methionine sulfone peaks. These peaks were at t.imes noticeable but uncalculable. 1 from ERSCO. to paper and electrophoresis was performed at pH 3.5 as described above. Carboxypeptidase A and B were used to determine the COOHterminal residue according to the method of Amber (19). Twenty micrograms of the pure peptide were mixed with 3 pg of carboxypeptidase A and 5 pg of carboxypeptidase B in 1 ml of buffer and incubated for 1.5 hours at 37". The released amino acids were estimated by electrophoresis at pH 1.9.

End Group Determinutions
Electrophoresis of an acid hydrolysate of the dansylpeptide revealed only dansyl-arginine.
However, when partially purified samples were subjected to the dansylation procedure, again only dansyl-arginine was found. Digestion with carboxypeptidase A and B released no amino acids from the pure peptide and did not destroy its sialogogic activity.
This indicates that the COOH-terminal residue is blocked or that at least 1 of the last 2 residues in the sequence is proline.

Enzymatic Studies
Trypsin-Incubation of the sialogogic peptide with trypsin did not destroy its activity nor did it cleave the peptide.
Electrophoresis at pH 3.5 of material incubated with trypsin again showed only one ninhydrin-positive spot; this spot corresponded to the only region of activity, and it had an electrophoretic mobilitjr identical with that of the untreated peptide.
Peps-in and Chymotrypsin-Incubation of the sialogogic peptide with pepsin or chymotrypsin completely destroyed the sialogogic activity. Incubation with inactivated pepsin or chymotrypsin had no effect. Table I summarizes the results of a typical preparative purification procedure starting with 20 kg of bovine hypothalami. The peptide had an electrophoretic mobility of 0.6 at pH 1.9 and of 0.7 at pH 3.5 (mobility: lysine, 1.0). Sialogogic activity was eluted from only one contiguous region of the paper at either pH. Fig. 5a shows the results of a typical electrophoretic run at pH 3.5. That electrophoresis at pH 3.5 yields a homogeneous peptide is indicated in Fig. 5b. When the material eluted from the horizontal strip of paper corresponding to the region numbered 3 on the stained guide strip (Fig. 5a) was rerun at pH 1.9, only one ninhydrin-positive spot was seen. However, since the amino acid compositions of material obtained after electrophoresis at pH 1.9 and 3.5 are the same (Table II), electrophoresis at either pH can serve as the final step in the purification procedure.

Amino Acid Composition
The amino acid composition of the pure peptide is given in Table II. The absorption spectrum of the pure peptide confirmed the lack of tyrosine and also showed the absence of trypto-  b Elevated glycine values were due to contamination during electrophoresis.
Blank strips of paper eluted as controls yielded glycine and smaller amounts of serine and alanine, respectively. The contamination problem was negligible when larger amounts of material were used.
c The low methionine value is partly explained by the fact that the methionine sulfone and methionine sulfoxide peaks were unusually high in this sample. When these peaks were calculated as part of the value for methionine, the molar ratio of methionine increased to at least 0.6.

Rot b !+$@j Duodenum
Guineo Pig   Ileum   I  I  1  I  ,  , , Electrophoresis at pH 6.6 and pH 8.9 In contrast to its behavior after electrophoresis for 30 min at pH 1.9 or at pH 3.5 where 90% of the sialogogic activity was recovered from a 2-to 3-cm strip of paper, the sialogogic peptide 8. Effects of pure sialogogic peptide on the blood pressure of an anesthetized rat. Body weight: 250 g. a, 0.2 pg per kg of peptide injected at arrow; 6, rat previously treated wit.h atropine, 50 mg per kg, 30 min before injection of 0.2 pg per kg of peptide at arrow.
after electrophoresis for 30 min was spread over 10 cm of paper at pH 8.9 and 15 cm of paper at pH 6.5. That this smear of activity can still be attributed to one peptide is supported by two observations.
(a) The homogeneous peptide obtained after electrophoresis at pH 3.5 was rerun at pH 6.5. Sialogogic material was eluted from two broad regions of the paper, but the amino acid compositions of the material from the two regions of paper were essentially the same (Table  III).
(5) Sialogogic material obtained after chromatography on Cm-cellulose was run first at pH 6.5. Samples eluted from the region of paper representing an electrophoretic mobility of 0.55 to 0.75 and of 0.2 to 0.55 (mobility: lysine, 1.00) were reapplied separately to paper and run under identical conditions. The faster moving material had the same broad distribution of activity as the original run, indicating that the smear of activity could be accounted for by one peptide. However, the electrophoretic mobility of the slower moving material was unchanged.
This strongly suggests that the sialogogic peptide is modified irreversibly during electrophoresis at 6.5.
At pH 8.9 the sialogogic activity still migrates toward the cathode with an electrophoretic mobility of 0.0 to 1.0 (mobility: lysine, 1 .O) .

Characterization of Sialogogic Peptide as Substance P Studies on Isolated
Intestinal Tissue-The pure peptide was found to stimulate the contraction of isolated guinea pig ileum and rat duodenum even in the presence of atropine, tryptamine, and Pyrilamine (Fig. 7). These results show that the activity of the peptide is not mediated via acetylcholine, serotonin, or histamine.
Blood Pressure Studies-Intravenous injection of one-tenth of a sialogogic dose of pure peptide lowered the arterial blood pressure of the anesthetized rat (Fig. 8). A nearly maximal fall in blood pressure was seen at one-fifth of a sialogogic dose. Preliminary treatment of the rat with atropine sulfate had no effect on the activity of the sialogogic peptide. DISCUSSION The sialogogic peptide isolated from bovine hypothalami is an undecapeptide containing Lysl, Arg,, Glxz, Proz, Gly,, Metr, Leur, and Phez. The calculated molecular weight of 1340 agrees well with the estimate of its size obtained by gel filtration (Fig.  6). The best evidence for the purity of the peptide reported here is the constant molar ratio of the constituent amino acids ob-Issue of September 25, 1970 M. M. Chang and S. E. Leeman  (Tables II and III).
Another strong piece of evidence is that when the undecapeptide obtained after electrophoresis at pH 3.5 was run at pH 1.9 only one ninhydrin-positive spot was seen and it corresponded to the only region of sialogogic activity (Fig. 5b). In this instance, the detection of a single NHz-terminal residue, arginine, is not evidence for purity, as partially purified samples also yielded only the danJy1 derivative of arginine after application of the dansyl procedure.
Two lines of evidence suggest that at least two of the three possible carboxyl groups in the peptide are not free: (a) a progressive increase in mobility with respect to lysine going from pH 1.9 to 8.9 electrophoresis, and (b) absorption on Cm-cellulose at pH 10.4.
Since Substance P has never been purified to homogeneity, it only has had a biological definition.
There is general agreement that at least one peptide referred to as Substance P has the following properties in the presence of atropine (2): (a) it is a potent hypotensive agent, (b) it stimulates the contraction of various smooth muscle preparations, and (c) it is distinguishable from kinins in that it can stimulate the contraction of rat duodenal tissue. The hypothalamic peptide isolated on the basis of its sialogogic activity meets these criteria.
Comparison of the chemical and physical properties of the sialogogic peptide with those reported for partially purified samples of Substance P from several different tissues and species suggests that these chemical entities are very similar if not identical (Table IV). Table V gives the amino acid compositions that have been reported for Substance P. The constituent amino acids in the sialogogic peptide are contained within the amino acid compositions reported for Substance P, with the exception of methionine.
It has been our experience and that of Zuber (7, 22) that the molar ratio of methionine relative to other amino acids increases with further purification.
Qualitatively, the amino acid compositions reported for Substance P contain Asp, Ser, Ala, Thr, Val, Ile, and Tyr, in addition to the amino acids found in the sialogogic peptide.
These same contaminating amino acids are found after amino acid analyses of our sialogogic preparations after Cm-cellulose chromatography but are separated from the sialogogic peptide by high voltage electrophoresis.
In further support of the close similarity of the sialogogic peptide and Substance P, Haefely and Hiirlimann (10) and Vogler et al. (9) noticed that their highly purified samples stimulated  The specific activity of the peptide isolated here appears to be at least 6 times greater than that reported for the most highly purified preparations of Substance P (6-10). Lembeck and Starke (5) reported that approximately 2,000 Euler units per Kg stimulated the secretion of 50 mg of saliva in rats. Assuming the sensitivity of their test rats to be the same as ours, one sialogogic dose would be the equivalent of 200 Euler units.
Thus the pure peptide has a calculated specific activity of 2,600,OOO Euler units per mg of protein.
The highest specific activity for a preparation of Substance P previously reported is that of Zuber (7) of 400,000 Euler units per mg of protein. Whether or not there is a single Substance P-peptide is not yet clear. Some preliminary evidence suggests that there may be more than one (8, 23). Zetler introduced a method for separating crude preparations of Substance P into three biologically active fractions, F,, Fb, and F,, by chromatography on aluminum oxide. Fraction F, is not Substance P as it does not have vasodepressor activity.
Both F, and Fb satisfy the original criteria established by von Euler and Gaddum (l), but F, produces relaxation instead of contraction of rat duodenal tissue. Recently, Baldauf and Gebhardt (24) have reported that a relaxing component can be separated from F, upon further purification, unmasking a peptide that can contract the rat duodenum. Whether the active substances in F, and Fb are peptides with different compositions or altered forms of the same peptide is not known.
Lembeck and Starke have shown that Fb can be converted to F, by shaking with aluminum oxide (25). The results of our electrophoretic experiments at pH 6.5 show that the sialogogic peptide can be changed into a slower moving component with no resultant change in amino acid composition.
Two other undecapeptides-physalaemin, isolated from the skin of a South American amphibian (26), and eledoisin, isolated from the salivary glands of cephalopods (27)-have been shown to have the biological a.ctivities of Substance P; i.e. they stimulate the contraction of smooth muscle tissue (5), and are potent vasodilators (28) and sialogogic agents (5). It is very interesting that the sialogogic peptide that we have isolated and characterized as Substance P is also an undecapeptide which has seven amino acids in common with physalaemin and eledoisin.
Furthermore, all three peptides stimulate the secretion of saliva at doses less than 2 pg per kg when given via tail vein in rats. The chemical and biological similarities of these three peptides make it tempting to speculate that these peptides had a common origin during evolution. AclcnowledgmentsWe wish to thank Mrs. Donna Fraser for her skillful technical assistance in all aspects of this work.
We also thank Dr. William Allison and Mr. Jan Admiraal for their helpful advice, Mrs. Lurley Cowell who performed the quantitative amino acid analyses, and Dr. Allen Herd, Department of Physiology, Harvard Medical School, in whose laboratory the blood pressure measurements were done.