The Isolation of a New Hypotensive Peptide, Neurotensin, from Bovine Hypothalami*

SUMMARY A hypotensive peptide, designated neurotensin, has been discovered and isolated in pure form from acid-acetone extracts of bovine hypothalami by column chromatography and paper electrophoresis. The results of amino acid analyses of material recovered after paper electrophoresis at pH 3.5, 6.5, and 8.9 and chromatographic analyses of the dansylated material indicate that the peptide isolated by this procedure is homogenous. Its amino acid composition and apparent molecular weight estimated by chromatography on Sephadex G-25 indicate that neurotensin is a tridecapeptide composed of Lys, Argz, Asx, Glxz, Proz, Ileu, Leuz, Tyr,.

From the Department of Physiology and Laboratory of Human Reproduction and Reproductive Biology, Harvard Medical School, Boston, Massachusetts 02115 SUMMARY A hypotensive peptide, designated neurotensin, has been discovered and isolated in pure form from acid-acetone extracts of bovine hypothalami by column chromatography and paper electrophoresis.
The results of amino acid analyses of material recovered after paper electrophoresis at pH 3.5, 6.5, and 8.9 and chromatographic analyses of the dansylated material indicate that the peptide isolated by this procedure is homogenous. Its amino acid composition and apparent molecular weight estimated by chromatography on Sephadex G-25 indicate that neurotensin is a tridecapeptide composed of Lys, Argz, Asx, Glxz, Proz, Ileu, Leuz, Tyr,.
Neurotensin lacks a free NH2-terminus; however, it possesses a free COOH terminus which can be acted upon by carboxypeptidase A. Neurotensin induces hypotension in the rat and can stimulate the contraction of guinea pig ileum and rat uterus; however, it produces relaxation of the rat duodenum.
These pharmacological properties classify it as a "kinin," yet its chemical composition distinguishes it from any known peptide.
During the course of purification of substance P from bovine hypothalamic extracts (l), the presence of a different peptide was detected that produces a visible vasodilation in the exposed cutaneous regions of anesthetized rats. This response occurs within seconds after intravenous injection and is associated with an acute hypotension.
Higher doses cause a distinctive cyanosis that persists for minutes.
This report presents a method for obtaining the peptide, which we have named neurotensin, in pure form, its amino acid composition, and some further characterization of its chemical and biological properties.
These data serve to distinguish it from other known mammalian peptides (2). The tissue fragments, extending from the optic chiasma to the mammillary bodies, weighed 8 to 10 g each. Albino rats of various sizes, hypophysectomized rats, and female guinea pigs weighing about 300 g were obtained from Charles River Breeding Labs. Sephadex media were obtained from Pharmacia Chemical Co., Silica Gel G plates from Schwartz-Mann, and chromatography papers from Whatman. Trypsin (25 units per mg), treated with L-(I-tosylamido-2-phenyl)ethyl chloromethyl ketone, and ol-chymotrypsin ( Depart,ment, Brandeis University, and w-as found by us to have a specific activity of 16 units per mg using benzoyl arginine methyl ester as substrate (3). Pronase (1 unit per mg) was a product of Sigma Chemical Co. Oxidized insulin A chain (glycyl) and the S-amino acid peptide, Phe-Val-Gln-Trp-Leu-Met-Asp-Thr, were products of Mann Research Labs. Synthetic oxytocin, synthetic vasopressin, synthetic physalaemin, and bacitracin were obtained from Sigma Chemical Co.

Analytical Procedures
Gel Chromatography and Ion Exchange Chromatography-The Sephadex G-25 and sulfoethyl-Sephadex C-25 were prepared as recommended by the manufacturer.
Descending gel chromatography was carried out at a flow rate of 0.3 to 0.5 ml per min per cm2 at room temperature or 4". Linear gradient elation of ion exchange columns was performed as described by Bailey (4). Column eluates were monitored by measuring their absorbance at 280 or 300 nm with a Zeiss spectrophotometer. Protein concentrations of column eluates were expressed as units of absorbance at 280 nm with the assumption that one absorbance unit represented 1 mg of protein per ml.
High Voltage Paper Electrophoresis-High voltage paper electrophoresis was performed on Whatman No. 1 or 3MM paper using a Michl type electrophoresis apparatus (5) at 80 volts per cm with Varsol as an inert cooling medium.
The buffer systems consisted of (a) pH 1.9, formic acid-acetic acid-water (150 : 100 : 750) ; (b) pH 3.5, pyridine-acetic acid-water (4:40:760) ; (c) pH 6.5, pyridine-acetic acid-water (80:2.4:720); and (d) pH 8.9, 1% ammonium carbonate in water. When preparative electrophoresis was to be carried out, the Whatman papers were washed with 20% pyridine and then 20% acetic acid for several days before use. Samples were applied to the paper in 0 2 M acetic acid, 0.01 M mercaptoethanol so as not to exceed a load of 0.1 mg of protein per cm and 1.0 mg of protein per cm on Whatman No. 1 and 3MM paper, respectively.
Appropriate standard amino acids (10 nmoles each) were spotted on both sides of the samples and peptides were located by staining a guide strip with ninhydrin-cadmium acetate reagent (6) or with the chlorine, o-tolidine method (7). Strips (1.5 to 2.0 cm in width) of the unstained section of the paper were eluted with 1 to 2 ml of 0.2 M acetic acid, 0.01 M mercaptoethanol (8). After lyophilization samples were diluted into 0.85% NaCl solution for biological testings. The electrophoretic mobility of neurotensin was calculated relative to lysine according to the method of Offord (9).
Amino Acid Analysis-Amino acid analyses were done according to Spackman et al. (10) with the aid of a Beckman model 120C automatic amino acid analyzer equipped with a high sensitivity adapter (II), precision =t3% on a 2 to 10 nmole range. Samples of peptides (5 to 10 nmoles) were routinely hydrolyzed with 0.25 ml of constant boiling HCl in evacuated, sealed tubes at 109" for 20 to 40 hours. Although tyrosine losses during acid hydrolysis have been reported to be about 10yc in 24 hours (la), we have found that at very low concentrations tyrosine can be degraded at 5-fold this rate (Fig. 1). The destruction rate is enhanced when the sample is first eluted from paper and impeded in the presence of phenol. Therefore, to minimize destruction of tyrosine, 25 ~1 of 0.1 M phenol were added to some tubes (13). For those analyses where phenol was not present tyrosine values were calculated from a time curve of the destruction extrapolated to zero time (Fig. 1). In order to quantitate the amounts of cysteine, cystine, and methionine in the peptide, performic acid oxidation was performed as described by Hirs (14). The dried sample was oxidized for 12 hours at 0" with 0.25 ml of performic acid reagent, diluted with cold water, lyophilized, and acid hydrolyzed for amino acid analysis.
Spectral analyses of the pure peptide were performed on a Cary model 15 recording spectrophotometer to determine tyrosine and tryptophan content.
Dansylation Proce&re-DNSl-chloride was used to determine the NHz-terminal residue according to the procedure of Gray (15). The DNS peptide was hydrolyzed in constant boiling HCl at 105" for 4 and 20 hours as recommended by Gros et al. (16) and examined by two-dimensional chromatography on 5 X 5-cm sheets of polyamide using the four solvent systems described by Hartley (17) Enzymatic Procedurea-Enzymatic digestions were done by 1 The abbreviation used is: DNS, l-dimethylaminonaphthalene-5-sulfonyl. incubating 10 doses of pure peptide with a 1:50 molar ratio of enzyme at 38" for 4 hours in 0.1 ml of the appropriate buffer. The buffer systems were as follows: pronase, 0.05 Tris-HCl buffer, pH 7.6; papain, 0.20 M ammonium acetate, pH 5.6, 0.03 M mercaptoethanol; thermolysin, 0.20 M NH4HC03, pH 8.2; oc-chymotrypsin, trypsin, and carboxypeptidase A and B, 0.10 DI NH-HC03, pH 7.8; leucine aminopeptidase, 0.05 M sodium barbital, pH 8.5, 0.01 M MgC&; aminopeptidase M, 0.06 M sodium phosphate, pH 7.8. Control solutions consisted of active pepticle incubated with enzyme that had been boiled for 5 min. The reactions were terminated by addition of 3 drops of acetic acid; the samples were lyophilized, dissolved in 0.857, saline, and tested in rats for their vasodilatory and hypotensive activity. In the cases where enzymatically released free amino acids were determined, approximately 200 doses of pure neurotensin were used per digest and analyses were performed on the amino acid analyzer.
Biological Procedures-Testings for the sialogogic activity of substance P were done according to the procedure of Leeman and Hammerechlag (1). One sialogogic dose is that amount of materia! that stimulates the secretion of 50 f 10 ~1 of saliva when injected intravenously into an anesthetized 100-g rat. The biological activity of neurotensin was monitored by observing the characteristic vasodilation of the exposed cutaneous regions of anesthesized rats that occurs within seconds following intravenous injection of samples. The intensity and duration of the response was noted and a minimal active dose was determined. Since this vasodilation was found to be associated with a transient fall in blood pressure, a dose of peptide was defined as that amount of material per 100 g body weight which when given intravenously to an anesthesized rat causes a fall in blood pressure of 35 f 5 mm Hg. Systemic blood pressure was measured with a Hewlett-Packard recorder and pressure transducer (preamplifier 8805 B, recorder 7782 A, transducer 267 UC) following cannulation of the carotid artery in rats weighing 250 to 300 g and anesthetized with pentobarbital (50 mg per kg). Test samples were dissolved in 0.85uj, saline and administered through a cannula in the femoral vein. The effects of pretreatments with various drugs on the response to neurotensin was examined by by guest on March 24, 2020 http://www.jbc.org/ the following procedure, each drug being tested in a separate group of rats. The animal was first shown to be responsive to an agonist in a dose producing a deviation in blood pressure of 30 to 40 mm Hg; then, 40 min after administration of the appropriate antagonist, the effect of the agonist was shown to be completely inhibited.
Shortly afterward neurotensin was tested in the same rat. The drugs, atropine sulfate (35 mg per kg), propranolol (5 mg per kg), and phenoxybenzamine (10 mg per kg) were administered subcutaneously to prevent the blood pressure responses to the agonists, acetylcholine (0.5 pg per kg), isoproteronol (0.5 pg per kg), and epinephrine (1.0 pg per kg), respectively.
Acute bilateral adrenalectomy was performed under pentobarbital anesthesia. Hypophysectomized rats were used within 24 hours of operation.
The effect of neurotensin on vascular permeability was measured using Evans blue dye, a dye that readily binds to plasma proteins, particularly albumin. Groups of rats anesthetized with pentobarbital (40 mg per kg) were injected intravenously with 2 ml per kg of a 1 y0 solution of Evans blue dye in 0.85% saline. Five minutes later 0.85% saline or neurotensin in 0.85% saline was injected either intravenously or intradermally.
After intravenous injection, the effect of neurotensin on the skin color of test animals was visually observed and compared to that of the control animals.
After intradermal injections the patches of skin around the injection sites (each weighing approximately 300 mg) were cut out 30 min later and the dye that had infiltrated the tissue was extracted with formamide and quantitated spectrophotometrically (18). The effect of neurotensin on the contract,ility of freshly dissected sections of guinea pig ileum, rat duodenum, and rat uterus was examined.
Rat uteri were taken from virgin rats determined to be in proestrous by their vaginal smear (19). All tissues were suspended in a 40.ml bath maintained at 37" and aerated with a mixture of Or-CO? (95:5).
Uterine contractility was examined in modified Locke's solution (20), while all other tissues were bathed in Tyrode's solution (21). All muscles were allowed to equilibrate in the bath for 30 min before the experiments were started.
After each response the bath was rinsed two to three times before addition of the next sample. The effects of added drugs on the response to peptide was examined after a 20.min incubation period.
Isotonic contractions of the muscles under a tension of 1 g were registered on a kymograph (Harvard Apparatus Co.) with a frontal lever giving a 5-fold magnification. The responsiveness of each tissue to standard solutions of synthetic bradykinin and substance I' was always tested. Whole blood, obtained anaerobically, w-as assayed for ~02, pCOs, and pH at 37" in an Instrumentation Laboratory analyzer (models 213 and 313). The analyses were done by Dr. Earl Weiss, Department of Respiratory Diseases, The St. Vincent Hospital of Worcester, Mass. Female albino rats, 90 to 130 g, were anesthetized with pentobarbital intraperitoneally (45 mg per kg) 40 min before use. Rats were then injected intravenously via tail vein with 0.85% saline or with neurotensin in saline, and arterial blood was withdrawn from the abdominal aorta into heparinized syringes at various times after injection. Samples were kept in ice water until the analyses were done.

Procedure for Isolation of Neurotensin
Extraction of Tissues-The frozen tissue (usually 20 to 45 kg with an assumed specific gravity of 1) was homogenized to a uniform consistency with an equal volume of -20" acetone-1 N HCl (100 :3 v/v) in a Gilford Wood colloid mill; then 3 more volumes of this solvent were added and the suspension was stirred overnight at 4". The mixture was suction filtered through Whatman No. 31 paper on Buchner funnels and the filtrate set aside. The residue was resuspended in a volume of acetone-0.01 N HCl (80 :20 v/v) that was three times the original volume of the tissue and filtered as above.
The two filtrates were pooled. Repetitive petroleum ether extraction of the combined filtrates to remove lipids as well as acetone was performed as follows: the filtrate was mixed at 4' with one-third its volume of petroleum ether (b.p. 36 to 50.9"); the ether-acetone phase was discarded and the process repeated three to four times until the discarded phase was transparent.
The acetone-water phase was then evaporated under reduced pressure (water bath temperature of 3545") to remove acetone and the aqueous residue was finally lyophilized.
The extraction and initial fractionation steps on a preparative scale were performed at the New England Enzyme Center, Tufts University, School of Medicine, Boston, Massachusetts.
Chromatography on G-25 Sephadex-Fractionation of the extract was performed in two successive steps on a 20.liter and a 5-liter column of Sephadex G-25 (fine). The activity of neurotensin is masked at this stage; however, substance P, a sialogogic undecapeptide (22) also present in these tissues, occupies the same region and was used as a marker.
The lyophilized extract was taken up in 0.1 M acetic acid (600 ml/22 kg of hypothalami) and after its pH was adjusted to 4, the suspension was centrifuged at 10,000 X g for 20 min at 4". The supernatant was resuspended in solvent (400 ml/20 kg of hypothalami) and recentrifuged.
The combined supernatant fluid was then applied to a 20.liter column of Sephadex G-25 (fine) which was equilibrated with 0.1 M acetic acid at 4" (Fig. 2)  was measured by injecting aliquots of eluates into test rats. pyridine-acetate, pH 5.5, in the reservoir.
The protein concentration of the eluates was monitored at 280 and 300 nm (Fig.  4), and neurotensin activity (unmasked by this fractionation step) was located by the visual vasodilation method. The active region was pooled, lyophilized, and the doses of neurotensin present were quantitated using the rat blood pressure method.
Material pooled from two of these columns of sulfoethyl-Sephadex (representing 45 kg of tissue) was applied in 100 ml of 0.01 &f pyridine-acetate, pH 5.5, to a lo-ml column of sulfoethyl-Sephadex which was then developed with a linear gradient made from 1.0 liter of 0.01 M pyridine-acetate, pH 5.5, and 1.0 liter of 0.3 M pyridine-acetate, pH 5.5 (Fig. 5). Preparative Paper Electrophoresis-High voltage paper electrophoresis at pH 3.5 on Whatman No. 3MM paper was used as the final preparative step. After the second ion exchange column the active material from 45 kg of hypothalami (about 10 mg of protein) was applied in a lo-cm band to Whatman No. 3MM paper and subjected to electrophoresis. Fig. 6a shows the pattern of ninhydrin staining material obtained and the doses of neurotensin recovered from corresponding regions of the paper. Material recovered from Region G (50% of the activity) was found to be a pure peptide.
The remainder of the activity can be recovered as a pure peptide by further eleetrophoresis at either pH 6.5 or pH 8.9. Neurotensin has an  Fig. 3; column size, 2.2 X 15 cm (resin volume 50 ml); fraction size, 20mI; column buffer, 6.05 M pyridine-acetate, pH 3.1. Neurotensin and sialogogic activities, determined by injecting lyophilized aliquots dissolved in 0.85y0 saline into test rats, were eluted in Fractions 50 to 60 and 65 to 75, respectively (see insets).

Purijkation
and Yields Table I summarizes the results of a typical purification procedure starting with 45 kg of bovine hypothalami.
By this procedure the extracted peptide is purified approximately ZOO,OOOfold. Approximately 3 to 5 nmoles of pure neurotensin can be obtained per kg of wet tissue. Assuming a 75% yield through the initial gel chromatography steps, it can be estimated that there are about 20 nmoles of this peptide per kg of tissue in the initial extract.
Since re-extraction of the initial precipitate with acetone-O:01 M HCI (60:40 v/v) yields more of a vasodilatory peptide that probably is neurotensin, there may be as much as 35 nmoles of neurotensin present per kg of hypothalami.
Thus, on a molar basis the concentration of neurotensin in bovine hypothalami is only about 20% that of substance I? (22) and 40 y0 that of luteinizing hormone-releasing factor (23).

Homogeneity and Composition
Several lines of evidence show that the biologically active peptide obtained is homogeneous.
The material obtained after electrophoresis at pH 3.5 runs as a single peptide during electrophoresis at pH 6.5 (Fig. Bb), and when the isolated peptide is treated with DNS-chloride and chromatographed on Silica Gel G only one fluorescent peptide spot is visible having an RF = 0.68 using butanol-pyridine-acetic acid-water (15: 10 : 3 : 12) and an Ra = 0.52 using butanol-acetic acid-water (4: 1: 1). The molar ratios of the constituent amino acids are integral and remain constant after electrophoresis at pH 3.5, 6.5, and 8.9 (Table II).
---, an estimation of the pyridine concentration. Neurotensin activity, measured after injection of lyophilized aliquots of every third fraction into test rats, was eluted from the column in Fractions 80 to 100 (see upper inset). To remove pyridine, which itself absorbs at 280 nm, each of these fractions was lyophilized, then each residue was dissolved in 2.0 ml of 0.10 M acetic acid and the Azso nm determined (see lower inset).  b Protein was calculated from quantitative amino acid analyses. Neurotensin, 7000 doses, was applied across 10 cm at the origin; 150 doses were spotted on the guide strip. Leu indicates leucine c Calculated by assuming a 75% yield of neurotensin through spotted at origin. Conditions: 60 min at 80 volts per cm; ninhy-the two initial gel chromatography stages.
drin stain. b, stained paper after electrophoresis at pH 6.5 of 1 ag of material recovered from region G shown in a as well as the Each column represents a se parate preparation.
of tyrosine per mole of peptide (Table II), and showed the absence of tryptophan. The absence of cysteine (ine) and methionine was established by amino acid analysis of 4 nmoles of pure peptide that had been subjected to performic acid oxidation. The following molar ratios were obtained: cysteic acid, 0; methionine sulfone, 0; aspartic acid, 1.05; glutamic acid, 2.0; proline, 2.0; isoleucine, 0.8; and leucine, 1.8. Tyrosine, lysine, and arginine were not determined. Fig. 8 shows the elution position of neurotensin as measured by vasodilatory activity relative to a number of other peptides of known molecular weight. The apparent molecular weight of neurotensin obtained from this plot is 1600. The minimum molecular weight of the tridecapeptide calculated from its amino acid composition is 1673.

End Group Determination
Neurotensin was found to lack a free NH2 terminus. Chromatography of the acid hydrolysates of neurotensin treated with DNS-chloride yielded only 0-DNS-tyrosine and N'-DNS-lysine and no cu-DNS-amino acid derivatives, indicating that the (Yamino group of the NHz-terminal residue is blocked. Additional support for this conclusion is that enzymatic digests of 10 nmoles of peptide with leucine aminopeptidase and 4 nmoles of peptide with aminopeptidase M did not release any free amino acids and did not destroy biological activity. Incubation of 10 nmoles of pure peptide with carboxypeptidase A and I3 released stoichiometric amounts of leucine, isoleucine, and tyrosine and destroyed its biological activity, indicating that the COOH terminus is free. The value in parentheses is that of tyrosine determined spectrophotometrically assuming a molar extinction coefficient of 974 nm = 1.4 X 103.

Enzymatic Studies
Incubation of biologically active material with the relatively nonspecific endopeptidases, pronase and papain, as well as enzymes with a high substrate specificity such as trypsin, chymotrypsin, and thermolysin destroyed its ability to produce vasodilation and to lower blood pressure in rats. These re-0.6. 0.5 Each point indicates the average elution position from three separate experiments with each of the following peptides: (1) Phe-Val-Gln-Tyr-Leu-Met-Asp-Thr; (2) arginine vasopressin; (3) physalaemin; (4) bacitracin; (5) oxidized insulin A chain. The elution position of neurotensin (applied sample, specific activity = 500 doses per mg of protein) was determined from the vasodilatory response following intravenous injection of eluates into rats.
sults strongly suggest that the biologically active substance is, at least in part, a peptide and that the peptide contains basic (e.g. lysine, arginine), aromatic (e.g. tyrosine), and bulky aliphatic (e.g. isoleucine, leucine) residues. The amino acid composition of the pure peptide is consistent with these findings.

Biological Studies
Vasodilation and Cyanosis-Intravenous injection of neurotensin (>200 pmoles per kg) produces within 30 s a visible dila-6859 tion of the small vesseIs in the exposed cutaneous regions of anesthetized rats, particularly noticeable in the ears, the feet, and around the mouth.
Larger doses (>l.O nmoles per kg) evoke this response and then cause a visible cyanosis within minutes after injection.
The strength and duration of these effects is directly related to the dose of neurotensin injected. The cyanosis is not associated with a change in the partial pressure of oxygen (~0~) or carbon dioxide (pCOz) in arterial blood. When arterial blood was drawn from a control group of anesthetized rats (nine animals) 2 to 3 min after intravenous injection with 0.85% saline, the following values (mean & standard error) were obtained: pOZ, 82 f 2 mm Hg; pCOZ, 38 & 1 mm Hg; pH 7.39 =t 0.02. Arterial blood drawn from the experimental group (nine animals) 1 to 3 min after intravenous injection of 2 nmoles per kg of neurotensin gave the following results; ~02, 84 + 4 mm Hg; pCOn, 36 f 4 mm Hg; pH 7.36 f 0.03. Rat Blood Pressure-Neurotensin is a potent hypotensive agent in the anesthetized rat, the threshold intravenous dose for a measurable response being about 100 pmoles per kg. The magnitude of the hypotensive effect depends upon the starting level of blood pressure, being diminished in rats with lower basal levels ( Table III).
The hypotensive effect of neurotensin exhibits acute tachyphylaxis, i.e. a second equal dose administered 1, 10, or 60 min after an active first dose produces no effect (Fig.  9); however, a second dose given several hours later is effective. 9. Effect of pure neurotensin on the systemic blood pressure of anesthetized rats and the tachyphylaxis that ensues. On the ordinate, carotid blood pressure in millimeters of Hg and on the abscissa, time in seconds. At each arrow 600 pmoles per kg were injected via femoral vein. Tracings a and b from one rat; tracings c and d from a second rat. neurotensin is given intradermally a greater amount of extractable dye accumulates at its site of the injection than that of saline (Fig. 10). Another indication of this increased permeability is the finding that the hematocrit of arterial blood increased from a control value (mean f range) of 46 f 2% in saline-injected animals (five rats) to 79 + 9% in the experimental group (seven rats) 15 min after intravenous injection of 0.2 nmoles per kg of neurotensin.

Studies on Isolated Intestinal
Tissue-Neurotensin stimulates the contraction of guinea pig ileum and rat uterus and the relaxation of rat duodenum in an organ bath (Fig. 11). Neurotensin was found to be as potent as bradykinin in its effect on guinea pig ileum and rat duodenum, but only one-fifth as potent as bradykinin on rat uterus. These responses still occur in the presence of atropine, tryptamine, pyrilamine, phenoxybenzamine, and propranolol, showing that these effects are not mediated by acetylcholine, serotonin, or histamine nor do they involve o(-or P-adrenergic receptors. DISCUSSION A new hypotensive peptide, designated neurotensin, has been isolated from bovine hypothalami and found to be a tridecapeptide composed of Lys, Argz, Asx, Glxz, Pro2, Ileu, Leuz, Tyrz. At the time of our initial report (24) the results indicated the presence of only one tyrosine moiety in this peptide. This is now known to have been an error caused by the degradation of tyrosine during acid hydrolysis of the minute amounts of peptide obtained.
The apparent molecular weight of the extracted biologically active material measured by gel chromatography is 1600; this is in good agreement with the minimum molecular weight of the isolated peptide as calculated from its amino acid composition.
However, since polypeptides can be generated from proteins during harsh stages of extraction and purification, it is conceivable that this tridecapeptide exists in hypothalamic tissue as part of a larger molecule.
The evidence indicates that although the COOH terminus of isolated neurotensin is free, the NH2 terminus is blocked.
It is possible that one of the two glutamic acid residues occupies the NH&erminal position in the cyclized pyrrolidone carboxylic acid form as is true of two other hypothalamic peptides, thyrotropin releasing factor and gonadotropin releasing factor (25).
Because neuroteusin is a hypotensive peptide that can contract rat uterus and guinea pig ileum and relax rat duodenum, it can be classified as a kinin (26). However, the cyanotic effect of neurotensin may be unique to this substance and may provide a simple means of pharmacological identification. Another distinguishing feature is the acute tachyphylaxis that is associated with the hypotensive action of this peptide. This property resembles that reported for the undecapeptide, ranatensin, of the blood pressure of the anesthetized dog (27). The potency of neurotensin in producing a change in vascular permeability is comparable to that reported for bradykinin (18). One can only speculate as to the physiological function of this peptide.
Since neurotensin is present in the hypothalamus, a reasonable suggestion is that it may play a neural role within this tissue perhaps as a transmitter or as a modulator of nervous activity.
It has been suggested that substance I', another peptidic neural constituent, may be a sensory transmitter (28) and, indeed, recent evidence indicates that this peptide at a concentration of lo-" M can cause depolarization of motor neurons in the spinal cord of the frog (29). However, the hypothalamus is also a major site of neurosecretion in mammals; therefore, an alternate suggestion is that neurotensin may function as neurohormone.
If this peptide does not prove to be a posterior pituitary hormone, nor a releasing factor for intermediate or anterior pituitary hormones, there is still an additional possibility. There is anatomical evidence to suggest that there are neurosecretory cells with short neurosecretory processes that synapse on capillaries within the hypothalamus (30). A neurosecretory substance could reach peripheral targets by this neurovascular pathway.
The biological properties of neurotensin that have been investigated are manifest following intravenous injection and are not eliminated by hypophysectomy of the test animal. If any of these responses operate physiologically it could represent an instance where a hypothalamic peptide plays a neurohormonal role affecting visceral function independent of the pituitary.
However, neurotensin may not be confined to the central nervous system; it may be a constituent of peripheral nerve or even other cells. The fact that neurotensin is a vasodilatory peptide which can induce changes in vascular permeability suggests that it may be one of the as yet unidentified peptides that play a role in neurogenic vasodilation (31) or the inflammatory response (32).