Design and Synthesis of Brain Penetrant Glycopeptide Analogues of PACAP With Neuroprotective Potential for Traumatic Brain Injury and Parkinsonism

Abstract

There is an unmet clinical need for curative therapies to treat neurodegenerative disorders. Most mainstay treatments currently on the market only alleviate specific symptoms and do not reverse disease progression. The Pituitary adenylate cyclase-activating polypeptide (PACAP), an endogenous neuropeptide hormone, has been extensively studied as a potential regenerative therapeutic. PACAP is widely distributed in the central nervous system (CNS) and exerts its neuroprotective and neurotrophic effects via the related Class B GPCRs PAC1, VPAC1, and VPAC2, at which the hormone shows roughly equal activity. Vasoactive intestinal peptide (VIP) also activates these receptors, and this close analogue of PACAP has also shown to promote neuronal survival in various animal models of acute and progressive neurodegenerative diseases. However, PACAP's poor pharmacokinetic profile (non-linear PK/PD), and more importantly its limited blood-brain barrier (BBB) permeability has hampered development of this peptide as a therapeutic. We have demonstrated that glycosylation of PACAP and related peptides promotes penetration of the BBB and improves PK properties while retaining efficacy and potency in the low nanomolar range at its target receptors. Furthermore, judicious structure-activity relationship (SAR) studies revealed key motifs that can be modulated to afford compounds with diverse selectivity profiles. Most importantly, we have demonstrated that select PACAP glycopeptide analogues (2LS80Mel and 2LS98Lac) exert potent neuroprotective effects and anti-inflammatory activity in animal models of traumatic brain injury and in a mild-toxin lesion model of Parkinson's disease, highlighting glycosylation as a viable strategy for converting endogenous peptides into robust and efficacious drug candidates.

Keywords: 6-hydroxydopamine; PAC1; VPAC1/VPAC2 receptors; blood-brain barrier; neurodegeneration; neuroinflammation.

FIGURE 1 |

Hypophysial hormones distribution to peripheral organs and brain by the circulatory system. (A) Numerous peptides (hypophysial hormones) are secreted by the pituitary gland, which are recirculated throughout the body, and which have effects in all of the organs, including brain. Effects within the brain are dependent upon BBB penetration rates. (B) The relevant endogenous hormones, PACAP_1–38, PACAP_1–27, and VIP_1–28 interact with 3 related Class B GPCRs, whose transmembrane structures are depicted as homology models in the inactive state.

FIGURE 2 |

PACAP glycopeptide drug design. We hypothesized that the introduction of a carbohydrate moiety at the C-terminal end of PACAP1–27 could improve stability and BBB permeability. We envisioned that replacement of Met¹⁷ with an alkyl side chain-containing amino acid and stabilizing substitutions at positions 1 and 2 would also enhance stability. Furthermore, we postulated that strategic amino acid substitutions at positions 4, 5, and 7 could lead to improved receptor selectivity profiles.

FIGURE 3 |

Hinge region comparison of PACAP and VIP. PACAP and VIP share a high primary amino acid sequence similarity but begin to diverge at positions 4 and 5. This relatively flexible “hinge” region allows the amphipathic helix to bind to the extracellular N-terminal domain of the Class B GPCR, and simultaneously interact with the transmembrane portion of the receptor. This region is sensitive to steric interactions between the 4^th and 5^th residues, and greatly influences selectivity of PAC1 and VPAC1 vs. VPAC2.

FIGURE 4 |

Synthetic scheme for preparation of PACAP-Derived glycopeptide.

FIGURE 5 |

Structures of non-natural amino acids used in these studies.

FIGURE 6 |

Additional modification of the Hinge region and C-terminus: 4^th Series of PACAP Glycopeptides. The identities of the carbohydrate residues at the C-terminus were re-investigated for their effects on potency, efficacy, and receptor selectivity. The hinge region was further explored for the effects of D-amino acids in the 4^th and 5^th positions on receptor selectivity, and a compound containing both Sar⁴ and Nle⁵ substitutions was also prepared and evaluated. Last, N-terminal acylation was further explored to determine how the acyl group affects receptor selectivity.

FIGURE 7 |

Summary of in vivo stability and BBB transport. (A) In vivo CSF concentrations of d82L98-OH and d82LS98Lact following 15 mg/kg i.v. injection in rats (n = 5). Glycopeptide concentrations were quantified by HPLC-MS. (B) Area under the curve (AUC) of d82L98-OH and d82LS98Lact following 15 mg/kg i.v. injection in rats. AUCs are normalized to d82L98-OH. d82L98-OH and d82LS98Lact are deuterated mass-shifted analogues of 2L98-OH and 2LS98Lact. Dosage of 15 mg/kg are used to overcome the high detection of limit in CSF. Glycosylated analogues exhibited enhanced BBB transportation with a disaccharide moiety. Concentrations and AUCs are reported by mean ± SEM. Asterisks demote statistical significance using repeated measures Mann-Whitney test, p < 0.01.

FIGURE 8 |

Summary of protective and anti-inflammatory effects of 2LS80Mel in a mouse model of TBI. Mice were subjected to diffuse TBI or control sham surgery and treated with either 2LS80Mel or sterile saline. The in vivo efficacy of 2LS80Mel was then evaluated by assessing its effects on the sleep-wake behavior (A–C), neurological and motor skill deficits (D–F), monocyte populations (G–I), concentrations of inflammatory cytokines (J–L), and microglial morphology (M–O). Overall, 2LS80Mel attenuated behavioral, neurological, and motor skill deficits. Furthermore, 2LS80Mel prevented increases in peripheral monocyte populations. There were no significant differences in inflammatory cytokine concentrations or microglial ramification between the 2LS80Mel-treated, untreated, and sham animals.

FIGURE 9 |

Evaluation of neuroprotection of the PACAP glycopeptide 2LS98Lac in rat Parkinson’s disease models. (A) Scheme for study 1, using a mild PD lesion. (B) Scheme for study 2, using a moderate PD lesion. (C) Systemic injection (i.p.) of 2LS98Lac reduces 6-OHDA-induced lesion damage in the mild 6-OHDA hemi-parkinsonian rat model. Amphetamine-induced rotations at 2- and 4-weeks post-lesion are plotted (mean rotations ± SEM). (D) The mean ± SEM cumulative amphetamine-induced rotations are plotted showing that treatment with the PACAP glycopeptide reduced the number of rotations indicative of a protective effect. *p < 0.05. (E) Unbiased stereology of TH-positive dopaminergic neurons in the substantia nigra (SNc) in study 1 reveal a significant 6-OHDA-induced loss of TH-positive neurons on the lesioned side in the vehicle control group (V), but not the 2LS98Lac-treated group (P). *p < 0.05. (F,G) Systemic injection (i.p.) of 2LS98Lac does not reduce 6-OHDA-induced lesion damage in the moderate 6-OHDA hemi-parkinsonian rat model. Amphetamine-induced rotations at 2 (F) and 4 (G) weeks post-lesion are plotted (mean rotations ± SEM). (H) Unbiased stereology of TH-positive neurons in the SNc in study 2 reveal a significant 6-OHDA-induced loss of TH-positive neurons on the lesioned side for both groups (V and P), but no group difference. ***p < 0.001. (I) 2LS98lac-treatment did not change mean (±SEM) striatal dopaminergic content (DA) analyzed with HPLC-EC. Mean data (±SEM) are plotted as % control. (J) 2LS98lac-treatment did not change striatal TH expression, quantified with semi-quantitative western analysis with beta-actin (βA) as internal standard. Mean data (±SEM) are plotted as % control. (K–N) 2LS98lac rescues 6-OHDA induced morphological changes to microglia (mean data ± SEM). Specifically, we show%area IBA1/cell (K), process length/cell (L), number of branches/cell (M) and endpoints/cell (N). *p < 0.05; **p < 0.01.

FIGURE 10 |

Example images for the analyses of the PACAP glycopeptide 2LS98Lac in the rat Parkinson’s disease model. (A) Example images of TH staining in the SNc that had been analyzed with unbiased stereology in Figure 9H to identify dopaminergic neurons within the SNc (40 μm thick serial sections were obtained and sampled every 480 μm. The scale bar is 500 μm. (B,C) Example western blots for the semi-quantitative western analysis done in Figure 9J for striatal TH with beta-actin (βA) as internal standard for the vehicle (B) and 2LS98Lac (C) groups. I: intact hemisphere; LX: lesioned hemisphere. (D) Example images for morphological microglia analyses, as done in Figures 9K–N, after confocal imaging of Iba1 stained brain sections (40 μm thick, sampled at 3 regions throughout the SNc). The scale bars are 10 μm.