Isolation and Characterization of a New Pancreatic Polypeptide Hormone*

A method is described for isolation, from chicken pancreas, of an avian pancreatic polypeptide which may be a new hormone. This method involves acid-alcohol extraction, gel filtration, DEAE-cellulose chromatography, and droplet countercurrent distribution. The peptide contains 36 amino acids, has a molecular weight of 4240 and the isoelectric point is pH 6 to ‘7. The average amount of avian pancreatic polypeptide extractable from chicken pancreas was 4 mg/lOO g of pancreas. The amino acid sequence of the peptide is Gly-Pro-Ser-Gln-Pro-Thr-Tyr-Pro-Gly-Asp-Asp-Ala-Pro-Val-Glu-Asp-Leu-Ile-Arg-Phe-Tyr-Asp-Asn-Leu-Gln-Gln-Tyr-Leu-Asn-Val-Val-Thr-Arg-His-Arg-Tyr-NH,. A previous report from this laboratory (1) described a peptide containing 36 amino acids which was isolated from chicken pancreas as a by-product of insulin purification. Additional studies have provided evidence that this avian pancreatic polypeptide may be a new pancreatic hormone. The evidence includes

A previous report from this laboratory (1) described a peptide containing 36 amino acids which was isolated from chicken pancreas as a by-product of insulin purification. Additional studies have provided evidence that this avian pancreatic polypeptide may be a new pancreatic hormone. The evidence includes demonstration by radioimmunoassay that the peptide could be localized to pancreas (2), a known endocrine organ, from which it could be extracted in amounts greater than or equal to the amount of extractable insulin. Recently it has been possible to demonstrate with immunofluorescence methods (3) that avian pancreatic polypeptide can be localized further to specific cells in the pancreas. No immunoreactivity could be detected in extracts of many other types of chicken tissue. In addition, immunoassayable avian pancreatic polypeptide was present in chicken plasma, and preliminary studies indicated that the plasma level increased in response to feeding (2). The peptide is not limited to chickens, since immunologically similar material was present in extracts of pancreas from a number of avian and reptilian species (2). However, it could not be demonstrated in extracts of amphibian or mammalian pancreas. This observation is of some interest since Lin and Chance (4) have actually isolated and determined the amino acid sequence of an obviously homologous peptide from bovine, porcine, ovine, and human pancreas. It has been shown (2) that the bovine peptide reacts only weakly, if at all, with antibody to avian pancreatic polypeptide.
Additional support for the hormonal character of avian pancreatic polypeptide and its bovine counterpart comes from to chickens, causes increased proventricular secretion of fluid, H+, and pepsin. In addition, from a metabolic standpoint, the peptide causes hepatic glycogenolysis without accompanying hyperglycemia. Lin and Chance and co-workers (4, 6-9) have observed effects of bovine pancreatic polypeptide upon a number of gastrointestinal functions in the dog. With regard to gastric acid secretion, bovine pancreatic polypeptide was capable of stimulation, but inhibited pentagastrin-induced secretion. In addition, basal and cholecystokinin plus secretin-induced pancreatic secretion was inhibited by bovine pancreatic polypeptide at doses as low as 1 hg/kg/hour. Gall bladder relaxation, increased choledochal tone, and decreased intestinal mobility have also been demonstrated in response to administered bovine pancreatic polypeptide. The pancreatic, gall bladder, and intestinal effects were not observed in chickens (5).
Since the original isolation of avian pancreatic polypeptide came about during chicken insulin isolation, and because this peptide is possibly a hormone, it was important to develop a more specific purification procedure. This report describes such a method and, in addition, the amino acid sequence of the peptide. The filtrate (Fraction 5A) was discarded. The SP-Sephadex to which APP and other peptides and proteins were bound was suspended in distilled water and allowed to swell overnight at 5". Following this, the slurry was brought to room temperature, and the Sephadex was packed into a column (4 x 50 cm). After passing 500 to 600 ml of water through this column, the bound peptides and protein were eluted with 0.4 M ammonium acetate adjusted to pH 9.5 with NH,OH. As elution proceeds a brown band forms and migrates down the gel column. Collection of eluate began with emergence of this band and continued for approximately 300 ml. The eluate was brought to pH 6 to 7 with acetic acid, concentrated to 50 to 70 ml at 40" in uacuo, and stored at -20" (Fraction 5  It may simply be a more acidic desamido-avian pancreatic polypeptide, or partially degraded avian pancreatic polypeptide, but it is also possible that this material is an avian pancreatic polypeptide precursor of higher molecular weight, such as is known to exist for insulin and glucagon (25-27).
As shown in Fig. 1 insulin was 65% based on immunoreactive material in the original extract.
Characterization of Product-The homogeneity of avian pancreatic polypeptide was verified by gel electrophoresis. A single band is visible at levels up to 100 &tube.
However, at this level a faint band which runs slightly anodic to the major avian pancreatic polypeptide band is visible.
It is possible that this material is desamido-avian pancreatic polypeptide.
Paper chromatography (1) of 100~ag samples of avian pancreatic polypeptide yields a single visible spot when stained with bromcresol green, and thin layer electrophoresis-chromatography shows only a single yellow ninhydrin-positive spot. (See Fig. 4). By radioimmunoassay, the glucagon content of the avian pancreatic polypeptide was 0.12% when assayed with antibody (30K) to porcine glucagon. Chicken glucagon and porcine glucagon react equally well with this antibody.6 The insulin content based on immunoassay using chicken insulin standards and antibody to chicken insulin was less than 0.001%. When injected into chickens (100 pg/kg) no detectable change in blood glucose was observed. Avian pancreatic polypeptide (10 mg/ml) yields a colorless solution; X,.. = 276 to 277 nm; h,,, = 248 nm; Ef, at 276 nm = 14.5.

Amino Acid Sequence Studies
Amino Acid Composition- Table  II gives the amino acid analyses of three different preparations of APP. Two of these are recent preparations which were purified by countercurrent distribution. The third analysis is from an earlier preparation in which final purification was accomplished by chromatography on CM-cellulose with urea-containing buffers.
It is noteworthy that all three preparations have the same amino acid composition, except for minor differences in the nonstoichiometric amounts of lysine present. None of the preparations contain detectable amounts of methionine or half-cystine, and tryptophan was shown to be absent both spectrophotometrically and by failure to react with Ehrlich's reagent.
No carbohydrate was detected with periodate (30) in any of the preparations.
On the basis of the amino acid composition the molecular weight of avian pancreatic polypeptide is 4240, and, since avian pancreatic polypeptide is retarded more than insulin on Sephadex G-50, this is most likely the correct molecular weight.  I  2  3  4 5 6  7 8  9 IO II I2 I3 I4 I5 I6 I7 18 I9  ?  777777 Gly-Pro-Ser-Gln-Pro-Thr-Tyr-Pro-Gly-Asp-A~-Ala-Pro-Vol-Glu-~p-Leu-Ile-Argf T  with hydrophobic side chains. The center segment (residues 17 to 31) consists of alternating hydrophobic polar pairs or singlets, while the COOH terminus is polar and highly basic. Only the center segment appears to be capable of forming a stable helical or p structure. The COOH terminus is too short, and the proline content of the NH,-terminal segment is too high. Thus, it is possible that avian pancreatic polypeptide exists as a coiled inner core from which two uncoiled peptide chains project. The longer NH,-terminal segment by virtue of its proline content might even be capable of folding back upon the core.
As mentioned earlier, Lin and Chance (8) have described a peptide from bovine pancreas that has an amino acid sequence similar to avian pancreatic polypeptide.
This homology is apparent when one compares the two sequences (Fig. 12)  is less than would be expected by chance. Homology suggests that the amino acid residues at positions 13 and 14 should be interchanged in either avian or bovine pancreatic polypeptide, but unless new information becomes available there is no basis for making a change at this time. It is of interest that bovine pancreatic polypeptide appears to preserve the primary structural features noted for avian pancreatic polypeptide; namely, the location of proline residues in the NH,-terminal segment; the alternating polar-hydrophobic side chains in the central portion and the basic COOH terminus.
There appears to be some homology between avian pancreatic polypeptide and chicken glucagon when the sequences are aligned as shown in Fig. 12. Eight identities occur and the average mutation value (34) is 1.14 (for random pairs the average mutation value is 1.45). It should be noted that both avian pancreatic polypeptide and glucagon are glycogenolytic (5), but have opposite effects upon gastric secretion. Lin and Chance (7,8)  in several avian and reptilian species (2) and homologous peptides have been isolated and characterized from bovine, ovine, porcine, and human pancreas (7, 8).
3. The pancreatic content of the peptides is of the same order of magnitude as that of insulin (2) and glucagon (33,35).
4. Avian pancreatic polypeptide and the human homolog have been localized to specific cells in the pancreas by immunofluorescent methods. In the case of avian pancreatic polypeptides, the cells are scattered throughout the pancreatic parenchyma (3), but in man they are located in the periphery of the islets (9). These cells are different from (Y, /3, and D cells, but have the ability to take up and decarboxylate L-Phe(OH), like the other pancreatic endocrine cells.
5. By radioimmunoassay, avian pancreatic polypeptide (36) or the appropriate homolog (8) has been demonstrated in the plasma of chickens, dogs, and humans. In the fasting state, blood levels are low, but they increase severalfold within 30 to 60 min after feeding. Thus blood levels change in response to physiological stimuli.
In chickens, partial pancreatectomy causes a significant decrease in plasma avian pancreatic polypeptide level and in the magnitude of the response to feeding (36). No recognizable deficiency state has been observed to date.
6. Avian and bovine pancreatic polypeptide have biological activity and are capable of eliciting a physiological response when administered in microgram amounts. One such response is gastrin-like in that intravenous injection of avian pancreatic polypeptide into chickens causes a marked increase in proventricular secretion of acid, fluid, and pepsin (5). In addition, avian pancreatic polypeptide is as potent as glucagon in causing hepatic glycogenolysis in chickens, but paradoxically this is not accompanied by hyperglycemia (5). The studies with bovine pancreatic polypeptide have been primarily concerned with effects upon gastric and pancreatic secretion, intestinal motility, and gall bladder activity in dogs. The peptide affects all of these activities in microgram amounts or less (4, 6-8).
In view of the above evidence for the hormonal nature of avian pancreatic polypeptide and the related mammalian peptides it is of importance to name the hormone before further proliferation of species specific abbreviations occurs. It is suggested that the name pancreatic hormone III be used to describe these peptides. This term names the source organ but has no functional connotation. The peptides represent the third hormone ascribed to the pancreas. Other known peptides such as gastrin (37), vasoactive intestinal peptide (38), or somatostatin (39) may originate in the pancreas but their secretion by that organ is not yet as well established as for APP. The use of Roman numeral III is suggested to prevent confusion with the term pH when pancreatic hormone III is abbreviated to PH-III. When more is known of the function of these peptides it may be desirable to devise a name which describes this function.