High molecular weight PEGylation of human pancreatic polypeptide at position 22 improves stability and reduces food intake in mice

Abstract

Background and purpose: Human pancreatic polypeptide (hPP) is known to suppress appetite and food intake, thereby representing a potential therapeutic approach against obesity and associated metabolic disorders. The aim of this study was to improve hPP stability by covalent PEGylation with diverse molecular weight polyethylene glycols (PEGs) at two positions using promising lead structures while maintaining target activity.

Experimental approach: Modified peptides were synthesized by combined solid-phase and solution-phase peptide synthesis. Their potency was investigated in constitutively expressing human epithelial cells and isolated human colonic mucosa as well as receptor-transfected artificial cell lines. Human blood plasma and porcine liver homogenates were used to examine the in vitro stability of the analogues. The most promising variants were injected s.c. in C57BL/6JRj mice to monitor fasting-induced food intake and bioavailability.

Key results: In human epithelia and colonic mucosal preparations, activity of the modified hPP peptides depended on the core sequence and latency of the peptides was related to PEG size. Peptides modified with a 22 kDa PEG (PEG22) remained intact in blood plasma and on incubation with liver homogenates for more than 96 h. Finally, hPP_2-36 , [K²² (PEG22)]hPP_2-36 and [K²² (PEG22),Q³⁴ ]hPP significantly reduced cumulative food intake in mice over 16 h after s.c. administration.

Conclusions and implications: Modification with PEG22 at position 22 stabilizes hPP significantly while extending its biological activities and could be used in drug development prospectively.

Figure 1

(A) Sequences of peptides indicating the position and type of PEG modification. For compounds 6a, 6b and 6c; see also (Mäde et al., 2014a). Analytical RP‐HPL chromatogram (B) and MALDI‐ToF mass spectrum (C) of purified [K²²(PEG22)]hPP_2–36 (4b), both illustrating broad peaks evoked by the PEG‐polydispersity. (B) A linear gradient of 20–60% CH₃CN (0.08% (v/v) TFA) in H₂O (0.1% (v/v) TFA) in 40 min with a Phenomenex‐Jupiter 4u Proteo column (90 Å, 4 μm, 250 × 4.6 mm) was applied. (C) MALDI‐ToF MS was performed in linear mode (M_{calcd .}: 26 170 Da, averaged M_obs. 25 779 Da).

Figure 2

Time‐dependent antisecretory responses of peptide analogues in Col‐24 monolayers (A, C) and human colon mucosa (B, D). Reductions in I_sc were recorded over 45 min in response to 100 nM of PEGylated [K²²]hPP_2–36 (A, B) or [K³⁰,Q³⁴]hPP variants (C, D) and predecessor molecules in both preparations. Data are the mean ± SEM (n numbers from different filters in A and C; and different tissue specimens in B and D) are shown in square brackets for each peptide.

Figure 3

Degradation of hPP analogues PEGylated at amino acid position 22 and respective precursor peptides. Peptide solutions (10 μM) were incubated with (A) human blood plasma or (B) 50 mg·mL⁻¹ porcine liver homogenates for the indicated times. Peptide‐related fluorescence was measured by RP‐HPLC at different time points and related to control samples at 0 h (100%). Data shown are the means of two independently performed experiments; individual values ranged between 99.9 and 116% of the mean. Results of the plasma degradation of [K²²,Q³⁴]hPP‐related peptides have been published (Mäde et al., 2014a).

Figure 4

Effects of PEGylated hPP analogues and precursor peptides on fasting induced food intake (A) and bioavailability (B) in C57BL/6JRj mice. (A) At the beginning of the dark phase, 18 h fasted mice were administered either with vehicle (saline; s.c.) or 3–20 mg·kg⁻¹ peptide s.c. dependent on the compound molecular weight. Progression of food intake by hY₄ receptor ‐targeting and hY₂ /hY₄ receptor ‐dual analogues for 48 h. Significances were calculated by ANOVA with Dunnett's multiple comparison test. The data significantly different (P ≤ 0.05) from vehicle are indicated as follows : * all peptides at 5 h: ~peptides 3, 4b, 6c at 10 h; # peptides 3, 4b at 15–20 h; § peptide 4b at 20–40 h. The arrow at 30h indicates a significant difference between peptides 4b and 3. Each line is the mean ± SEM (n ₃ = 5 and 5, meaning 5 per run, n _4b = 5 and 5, n _6a = 5 and 6, n _6b = 4 and 6, n _6c = 5 and 5, n _vehicle = 6 and 6; randomized by block randomization). Please note the error bars have been omitted in this data set for the sake of clarity. (B) Bioavailability in mice was markedly elevated for 4b, while 3 was cleared from the body within 6 h. Data shown are means ± SEM of samples from five (peptides 3, 4b) and means of three (vehicle), different animals per data point; individual values from vehicle treated animals ranged from 28 to 243 % of the mean.