Dipeptidyl peptidase IV (DPP-IV) from pig kidney cleaves analogs of bovine growth hormone-releasing factor (bGRF) modified at position 2 with Ser, Thr or Val. Extended DPP-IV substrate specificity?

The literature reported DPP-IV substrate specificity includes oligopeptides with a penultimate Pro, Hyp or Ala residue. Bovine GRF is a substrate for DPP-IV and is rapidly degraded by the enzyme via removal of its N-terminal Tyr-Ala. Incubation of selected GRF analogs from the [X2,Ala15,Leu27]bGRF(1-29)NH2 series with a porcine-kidney-derived DPP-IV in PBS (pH 7.4) resulted in cleavage at the X2-Asp3 bond. The extent of enzymatic hydrolysis varied with X2 as reflected in the following relative cleavage rates: Ala2 (100%), Ser2 (4%), Thr2 (2.5%), Val2 (0.53%), Ile2 (0%). These cleavages were sequestered when similar experiments were performed in the presence of the DPP-IV-specific inhibitor N-epsilon-(p-NO2-benzyloxycarbonyl)-Lys-Pro-OH. A side reaction, buffer-induced deamidation of Asn8, contributed less than 5% of the total substrate degradation. Although our finding qualitatively extends the DPP-IV substrate specificity to also include N-terminal X-Ser, X-Thr and X-Val sequences, quantitatively, relatively fast cleavages of the GRFs with Ala2 make the latter preferred substrates for DPP-IV. The data presented here indicates that the observed GRF(3-29) fragment formation upon incubation of Ser2- and Thr2-substituted bGRF analogs in bovine plasma could have been DPP-IV-related.

Position 2 and position 2/Ala15-substituted analogs of bovine growth hormone-releasing factor (bGRF) with enhanced metabolic stability and improved in vivo bioactivity.

In order to prepare GRF analogs with high activity in vivo, a strategy was undertaken to stabilize the peptide to dipeptidylpeptidase IV (DPP-IV), a protease found in plasma which inactivates native human and bovine GRF by cleavage of the Ala2-Asp3 bond. Replacement of the Ala2 residue with Ser, Thr, or Gly in [Leu27]bGRF(1-29)NH2 resulted in peptides greatly stabilized against proteolysis in plasma, but having low inherent GH-releasing activity when tested in bovine pituitary cell cultures. Replacement of Gly15 with Ala15 was marginally effective in improving the in vitro bioactivity of this group of peptides. When tested for GH-hormone release in steers, however, the Thr2,Ala15 analog was four times more potent than bGRF(1-44)NH2. Eleven additional analogs from the [X2,Ala15,Leu27]bGRF(1-29)NH2 series were synthesized and evaluated for metabolic stability in bovine plasma and for GH releasing activity in steers in vivo and in rat pituitary cells in vitro. Two compounds, [Val2,Ala15,Leu27]dGRF(1-29)NH2 and [Ile2,Ala15,Leu27]-bGRF(1-29)NH2, had increased GH-releasing activity in steers over that of [Thr2,Ala15,Leu27]-bGRF(1-29)NH2 and over a previously reported super-potent analog, [desNH2Tyr1,D-Ala2,Ala15]-hGRF(1-29)NH2.

Growth hormone-releasing factor analogs with hydrophobic residues at position 19. Effects on growth hormone releasing activity in vitro and in vivo, stability in blood plasma in vitro, and secondary structure.

To test the hypothesis that replacing Ala19 in growth hormone-releasing factor (GRF) with more hydrophobic residues will increase growth hormone releasing activity, four GRF analogs were prepared and tested. The molecules were made by substituting Val, Ile, or Leu at position 19 of [Thr2,Ala15,Leu27]bGRF(1-29)NH2. The compounds were evaluated for growth hormone (GH) releasing activity in vitro (rat anterior pituitary cells) and in vivo (steers). Additionally, their half-life in vitro was determined in bovine plasma, and their secondary structure was examined by circular dichroism. In pituitary cells, peptides with substitutions at position 19 had the following potencies: Ala (native), 0.37; Val, 1.16; Ile, 0.37; Leu, 0.043. When assayed in steers as a single iv bolus, over a 2-h period, the compounds gave the following integrated GH response: Ala, 2.75; Val, 2.67; Ile, 2.57; Leu, 1.55. Only the Leu analog was statistically different from the other three (p = 0.05). In bovine plasma, the half-lives (hours) were as follows: Ala, 4.9; Val, 6.6; Ile, 12.3; Leu, 14.7. In phosphate buffer the compounds were calculated to have the following percent helical content: Ala, 26; Val, 21; Ile, 27; Leu, 32. For these analogs, helicity in aqueous buffer is inversely related to their in vitro activity. Using a linear multiple regression model, the plasma half-life of the analogs positively correlated (r2 = 0.999) with both the hydrophobicity of the residue at position 19 and the helicity of the analog. Although the Val analog had both increased inherent activity and increased plasma stability in vitro compared to the Ala analog, in this study we were unable to demonstrate an increase in activity in vivo. The in vivo GH releasing activity of the analogs was not simply related to a combination of their intrinsic GH releasing activity and their in vitro plasma half-life. This suggests that in vivo additional factors are moderating the expression of activity.

Metabolism of mouse growth hormone-releasing factor, mGRF(1-42)OH, and selected analogs from the bovine GRF series in mouse and bovine plasma in vitro.

The presence of Val2 in mGRF(1-42)OH is unique and, as shown in this study, renders this GRF resistant to plasma DPP-IV, the main enzyme responsible for rapid hydrolysis and inactivation of Ala2-containing GRFs from other species via cleavages between Ala2-Asp3. The presence of DPP-IV activity in mouse serum, and mouse and bovine plasma has been demonstrated with Gly-Pro-p-nitroanilide and/or with two DPP-IV-sensitive bGRF analogs, [Leu27]bGRF(1-29)NH2 and [Ala15,Leu27]bGRF(1-29)NH2, which were effectively converted to their respective (3-29) fragments. During incubations of mGRF(1-42)OH in mouse serum or plasma, as well as in bovine plasma in vitro, no major fragments were detectable, except for small amounts of metabolites with HPLC retention times corresponding to those of mGRF(12-42)OH and mGRF(21-42)OH, indicative of possible trypsin-like cleavages between Arg11-Lys12 and Arg20-Lys21. Both mGRF(1-42)OH (t1/2 52-78.5 min) and [Val2,Ala15,Leu27]-bGRF(1-29)NH2 (t1/2 78.5 min) disappeared 5 to 7 times faster in mouse than in bovine plasma, indicating much higher activity of various degrading enzymes in mouse plasma. In summary, our data provide evidence that mGRF(1-42)OH, despite its resistance to plasma DPP-IV, is degraded relatively fast in mouse plasma or serum because of trypsin-like and other, non-DPP-IV-related, proteolytic cleavages.

1H NMR analysis and in vitro bioactivity of Leu27-bGRF(1-29)NH2 and its D-Ala2 and des-(Tyr1-Ala2)-analogs.

Relative growth hormone-releasing potencies of bovine growth hormone-releasing factor (bGRF) analogs bGRF(1-44)NH2 (I), Leu27-bGRF(1-29)NH2 (II) and D-Ala2, Leu27-bGRF(1-29)NH2 (III) in in vitro bovine anterior pituitary cell cultures were determined to be 100%, 48% and 77%, respectively. The potencies of II and III, although numerically different, were not statistically different. Leu27-bGRF(3-29)NH2 (IV) was approximately 10,000 times less potent than 1. 1H NMR studies of peptides II, III and IV in 35% d3-2,2,2-trifluorethanol (TFE)/65% phosphate buffer at pH 4 revealed very similar, highly helical secondary structures in the 8-29 region, with only subtle differences at the N-termini. This lack of correlation between secondary structure in solution and in vitro bioactivity suggests that either 1) the biological conformations induced at the GRF receptor for II and III vs. IV are different from those generated in TFE/buffer, 2) similar secondary structures may be necessary but not sufficient for the observed bioactivity or 3) residues 1 and 2 of analogs II and III are important contact residues crucial for effective GRF-receptor interaction.