Research Article Cite This: ACS Chem. Neurosci. XXXX, XXX, XXX-XXX
pubs.acs.org/chemneuro
Positron Emission Tomography Assessment of the Intranasal Delivery Route for Orexin A Genevieve C. Van de Bittner,†,∇ Kyle C. Van de Bittner,†,○ Hsiao-Ying Wey,† Wayne Rowe,‡ Ram Dharanipragada,‡ Xiaoyou Ying,∥ William Hurst,‡ Andrew Giovanni,‡ Kim Alving,⊥ Anurag Gupta,∥,# John Hoekman,§ and Jacob M. Hooker*,† †
Department of Radiology, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts 02129, United States ‡ Sanofi US, Bridgewater, New Jersey 08807, United States ∥ Sanofi US, Framingham, Massachusetts 01701, United States ⊥ Sanofi US, Waltham, Massachusetts 02451, United States § Impel NeuroPharma, Seattle, Washington 98119, United States S Supporting Information *
ABSTRACT: Intranasal drug delivery is a noninvasive drug delivery route that can enhance systemic delivery of therapeutics with poor oral bioavailability by exploiting the rich microvasculature within the nasal cavity. The intranasal delivery route has also been targeted as a method for improved brain uptake of neurotherapeutics, with a goal of harnessing putative, direct nose-to-brain pathways. Studies in rodents, nonhuman primates, and humans have pointed to the efficacy of intranasally delivered neurotherapeutics, while radiolabeling studies have analyzed brain uptake following intranasal administration. In the present study, we employed carbon-11 radioactive methylation to assess the pharmacokinetic mechanism of intranasal delivery of Orexin A, a native neuropeptide and prospective antinarcoleptic drug that binds the orexin receptor 1. Using physicochemical and pharmacological analysis, we identified the methylation sites and confirmed the structure and function of methylated Orexin A (CH3-Orexin A) prior to monitoring its brain uptake following intranasal administration in rodent and nonhuman primate. Through positron emission tomography (PET) imaging of [11C]CH3Orexin A, we determined that the brain exposure to Orexin A is poor after intranasal administration. Additional ex vivo analysis of brain uptake using [125I]Orexin A indicated intranasal administration of Orexin A affords similar brain uptake when compared to intravenous administration across most brain regions, with possible increased brain uptake localized to the olfactory bulbs. KEYWORDS: Orexin A, hypocretin-1, intranasal, positron emission tomography, pharmacokinetics, peptide radiolabeling
■
perivascular channels of the lamina propria.2,4,5 In addition, slower (i.e., hours to days) intracellular transport through the olfactory sensory neurons (OSNs) has been suggested, which requires intracellular uptake into the OSNs, slow transport along the axon, and exocytosis into the brain.4,6−8 Transport to the brain may also be assisted by injury to the nasal neuroepithelium caused by intranasal injection of a therapeutic, particularly for larger injection volumes, >100 μL/nare for humans9 or >10−25 μL/nare for rats.3,10 Intranasal delivery has been recently applied to the neuropeptide Orexin A, which is expressed in the hypothalamus11−13 and is important for regulating energy homeostasis,14 sleep−wake cycles,12,15−17 and the endocrine system.12,13 Early studies in rodent indicated that exogenous
INTRODUCTION Intranasal drug administration has been explored as a unique method to deliver neurotherapeutics more directly to the brain while potentially limiting unwanted drug exposure to systemic organs. Improved delivery to the brain following intranasal administration can be achieved through several enrichment processes, namely, bypassing metabolic breakdown in the stomach and/or liver, avoiding absorption limitations at the gut wall, and circumventing the blood-brain barrier (BBB).1,2 For peptide neurotherapeutics, the advantages may be especially pronounced, as this class of molecules exhibits highly facile amino acid bond cleavage within the digestive system and a decreased BBB penetrance compared to small molecule neurotherapeutics.3,4 The increased CNS penetration and therapeutic efficacy demonstrated following intranasal administration is often attributed to relatively fast (i.e., minutes to 1 h) extracellular transport along either the perineural space surrounding the olfactory and trigeminal nerves or along the © XXXX American Chemical Society
Received: September 12, 2017 Accepted: October 16, 2017 Published: October 16, 2017 A
DOI: 10.1021/acschemneuro.7b00357 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
Research Article
ACS Chemical Neuroscience
Figure 1. CH3-Orexin A synthesis and methylation site identification. (A) General reaction scheme for CH3-Orexin A synthesis, i: CH3I, KOH (s), DMSO, RT, 5 min. (B) Peptide sequence for Orexin A with indication of most probable methylation sites (10Lys and 17Tyr) based on LCMS data. (C) MS data analysis confirms Orexin A methylation. The percent methylated product is kept low for identification of the most reactive nucleophile site(s) with the Orexin A peptide and to approximate carbon-11 radiolabeling conditions. (D) MS2 of the intact M4 and M5 ions reveals fragments that localize methylation to amino acid residues 1−29. (E) Further fragmentation using MS3 provides data supporting amino acid labeling within residues 1−19.
of the administered dose. A study utilizing intranasal administration of [125I]Orexin A was completed in rats by Dhuria et al.16 Gamma counting and autoradiography were used to assess the distribution of [125I]Orexin A in rat brain and the results were compared to intravenous administration. A modest uptake of Orexin A into the brain after both intravenous and intranasal administration was detected, with a similar distribution pattern across brain regions and highest uptake within the trigeminal nerve and the olfactory bulbs. Importantly, there was lower systemic uptake of [125I]Orexin A following intranasal administration. Although this study suggests that intranasally administered Orexin A reaches the brain of rodents in low doses within 1 h,16,22 these studies have not confirmed the precise location of the [125I]Orexin A dose after intranasal administration while also monitoring subsequent brain uptake in a temporally resolved fashion. Completing an analysis that can correlate dose localization with brain uptake with temporal fidelity could provide a deeper understanding of the mechanism of action for the intranasal delivery route. In this study, we utilized positron emission tomography (PET) to assess the level and route of brain uptake of Orexin A following intranasal administration in rats and NHPs. To achieve these goals, we developed a method to radiolabel Orexin A with [11C]methyl iodide to form [11C]CH3-Orexin A. With [11C]CH3-Orexin A in hand, a gel electrophoresis technique was developed to confirm the covalent, stable attachment of the radiolabel to Orexin A. After confirming CH3-Orexin A’s carbon-11 labeling site, its alpha helical peptide secondary structure, and its ability to bind and agonize orexin receptor 1 (OX1), PET imaging was used to assess the
Orexin A may be beneficial for the treatment of narcolepsy, as intracerebroventricular (ICV) administration of Orexin A improved wakefulness of mice with a diminished ability to produce Orexin A.18 While these results were encouraging, ICV administration is highly invasive and technically challenging, limiting its application as a dosing strategy in human patients. To overcome the obstacle of direct brain injection, more recent studies have assessed the therapeutic efficacy of intranasal Orexin A administration in sleep-deprived nonhuman primates (NHPs) and narcoleptic human patients.15,19 In NHPs, intranasal Orexin A was found to reverse the effects of sleep deprivation, specifically improving short-term memory performance and reversing alterations in cerebral glucose metabolism induced by sleep deprivation.15 The same study showed that intravenous administration with a 5−10 fold higher dose of Orexin A could also reverse the effects of sleep deprivation. In narcoleptic human patients, intranasal Orexin A was shown to have a sleep-stabilizing effect that led to improved performance on a divided-attention test; however, no wake-promoting effects were observed at the dose tested.19 There are several hypotheses for the mechanism of action for intranasally administered Orexin A, the favored hypothesis being direct transport of Orexin A from the nasal cavity to the brain. It is also possible that the physiological response is produced by the high, localized dose of Orexin A within the nasal cavity, which could bind orexin receptors presented by the OSNs,13,20,21 leading to a signaling cascade that propagates into the brain. To address the question of mechanism of action related to delivery pharmacokinetics (i.e., exposure), intranasal injection can be completed with radiolabeled molecules to allow tracing B
DOI: 10.1021/acschemneuro.7b00357 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
Research Article
ACS Chemical Neuroscience
Figure 2. Analysis of 10Lys and 17Tyr Orexin A methylation sites. (A) LCMS TIC of methylated Orexin A (left). Isotope pattern for unmethylated Orexin A (z = 3, right). (B) LCMS XIC of CH3-Orexin A (left). Isotope patterns (z = 3) of the 10Lys-CH3-Orexin (retention time (RT) = 5.8 min, right, upper panel) and 17Tyr-CH3-Orexin A (RT = 5.9 min, right, lower panel). (C) LCMS TIC of 10Lys-CH3-Orexin A standard (left) with the 10Lys-CH3-Orexin A isotope pattern (z = 3, right). (D) LCMS TIC of 17Tyr-CH3-Orexin A standard (left) with the 17Tyr-CH3-Orexin A isotope pattern (z = 3, right). The retention times for the three major peaks seen in the CH3-Orexin A XIC match the retention times for 10Lys-CH3-Orexin A and 17Tyr-CH3-Orexin A, indicating that CH3-Orexin A is a mixture of 10Lys and 17Tyr methylated peptides.
To identify the methylation site(s) of CH3-Orexin A, a liquid chromatography−mass spectrometry (LCMS) analysis was completed using MS2 and MS3 tandem mass spectrometry. This analysis indicated that the CH3-Orexin A methylation site was within the first 19 amino acids of Orexin A (Figure 1B−E). Given the amino acid side chains present within this region, the most likely positions of methylation were the phenolic oxygen of the tyrosine at position 17 (17Tyr) and the primary amine of the lysine at position 10 (10Lys) (Figure 1B). To assess these two possibilities, solid-phase synthesized standards of Orexin A methylated at either 10Lys or 17Tyr were obtained, and a liquid chromatography method was developed to separate these two CH3-Orexin A standards for assessment by mass spectrometry (Figure 2). Comparison of the 17Tyr-CH3-Orexin A and 10Lys-CH3-Orexin A standards to CH3-Orexin A indicated that CH3-Orexin A contained a mixture of Orexin A methylated at the 10Lys and 17Tyr residues (Figure 2). On whole, a successful method for methylation of Orexin A was developed and the methylation sites were determined. Assessment of the Secondary Structure, IC50, and EC50 of CH3-Orexin A. Confirmation of the secondary structure of CH3-Orexin A following the methylation reaction was provided via circular dichroism. The overlay of Orexin A and CH3-
intranasal delivery route. The excellent temporal resolution, noninvasive analysis, and high sensitivity provided by PET imaging offered a method to monitor the intranasally injected Orexin A dose in real time, throughout the nasal cavity and brain.
■
RESULTS AND DISCUSSION
Site-Specific Methylation of Orexin A. Toward the goal of radiolabeling Orexin A using the canonical [11C]methyl iodide ([11C]CH3I) labeling chemistry, we developed a nonradioactive methylation procedure that mimicked common radiolabeling conditions. Specifically, radiolabeling commonly occurs with substoichiometric quantities of the radiolabeling agent and requires very short reaction times, usually ≤5 min for carbon-11. To maintain the Orexin A peptide secondary structure, we focused our efforts on peptide-friendly reaction solvents, low temperatures, and limited levels of base for reaction catalysis. Ultimately, dissolving Orexin A in dimethyl sulfoxide (DMSO) and adding minimal solid potassium hydroxide (KOH) in the presence of CH3I with stirring for 5 min provided the singly methylated Orexin A product, CH3Orexin A (Figure 1A). C
DOI: 10.1021/acschemneuro.7b00357 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
Research Article
ACS Chemical Neuroscience
Analysis of the binding and functional agonism characteristics of CH3-Orexin A was accomplished using the synthesized 17Tyr-CH3-Orexin A and 10Lys-CH3-Orexin A standards, which were compared to unmodified Orexin A. All three compounds were tested for their binding affinity (IC50) and agonist activity (EC50) with human OX1. The IC50’s of 17TyrCH3-Orexin A (1.0 × 10−8 M) and 10Lys-CH3-Orexin A (2.7 × 10−8 M) with OX1 were slightly higher, but comparable to the IC50 of the parent Orexin A (7.7 × 10−9 M), indicating limited effect of Orexin A methylation on the binding interaction with OX1 (Figure 3B). The functional agonism results were similar, with the 17Tyr-CH3-Orexin A EC50 (6.3 × 10−9 M) being equivalent to the Orexin A EC50 (6.1 × 10−9 M), while 10LysCH3-Orexin A exhibited a reduced EC50 (
