Preliminary PET Imaging of Microtubule-Based PET Radioligand [¹¹C]MPC-6827 in a Nonhuman Primate Model of Alzheimer’s Disease

Abstract

The microtubule (MT) instability observed in Alzheimer’s disease (AD) is commonly attributed to hyperphosphorylation of the MT-associated protein, tau. In vivo PET imaging offers an opportunity to gain critical information about MT changes with the onset and development of AD and related dementia. We developed the first brain-penetrant MT PET ligand, [¹¹C]MPC-6827, and evaluated its in vivo imaging utility in vervet monkeys. Consistent with our previous in vitro cell uptake and in vivo rodent imaging experiments, [¹¹C]MPC-6827 uptake increased with MT destabilization. Radioactive uptake was inversely related to (cerebrospinal fluid) CSF Aβ₄₂ levels and directly related to age in a nonhuman primate (NHP) model of AD. Additionally, in vitro autoradiography studies also corroborated PET imaging results. Here, we report the preliminary results of PET imaging with [¹¹C]MPC-6827 in four female vervet monkeys with high or low CSF Aβ₄₂ levels, which have been shown to correlate with the Aβ plaque burden, similar to humans.

Graphical Abstract

1. INTRODUCTION

Microtubules (MTs) are important components of the cytoskeleton. Along with actin, MTs contribute to structural integrity of the cellular network.¹ They assist in direct molecular trafficking across subcellular compartments and process the information in key cellular signaling activities.² MT dynamics—changes in conformation, stability, length, and number—are highly regulated in healthy cells to perform physiological functions, such as cell division, axoplasmic transport, and cell signaling.^3,4 Their dysfunctions particularly affect critical mechanisms in brain cells, which leads to neuronal and cognitive abnormalities.^5–7 Abnormalities in MT-based neuronal structures and cellular functions are heavily implicated in several disorders, including Alzheimer’s disease (AD), substance use disorder, amyotrophic lateral sclerosis, multiple sclerosis, cancer, Parkinson’s disease (PD), and other psychiatric diseases.^8–13 Since tau, an MT-associated protein, is highly hyperphosphorylated in AD pathogenesis, MT dysregulation is intensely studied in pursuit of diagnostic and therapeutic interventions for AD and AD-related dementias (ADRD).^12,14–16

In healthy neurons, tau binds to MTs to regulate their stability. In AD brains, in a mechanism thought to be triggered by amyloid beta (Aβ) pathology, tau detaches from MTs and hyperphosphorylates, which results in pathological tau species that mediate neuron loss and cell death.¹⁷ Although clinical studies of emerging MT agents have shown promising results in brain malignancies and tauopathies, including AD, PD, progressive supranuclear palsy (PSP), and corticobasal degeneration, we do not know whether MT abnormalities have a causal and/or early role in the AD process or if they represent a common downstream end point in the neurodegenerative cascade. As the key event in AD is an increase in the Aβ levels that eventually initiates tau pathology,¹⁷ real-time imaging of changing molecular concentrations of MTs and comparison with Aβ₄₂ may answer these questions.^1,18–22

PET (positron emission tomography) is a highly sensitive nuclear imaging technique that uses radioactive ligands to detect and quantify biomarkers in vivo.^23–25 Functional PET imaging of MTs using specific radiotracers could help to define MT dysregulation to improve disease diagnosis and treatment monitoring and facilitate the development of novel therapeutics.^11,26,27 As Aβ and tau pathologies accumulate in the AD brain, MT integrity is compromised. Therefore, PET imaging of MT-based processes can provide an early indication of neuronal health and brain function prior to the onset of AD symptoms. MPC-6827 was identified as an MT-targeting agent (MTA) that binds to tubulin sites with high affinity (IC₅₀ = 1.5 nM).²⁸ MPC-6827 was used to treat glioblastoma in several clinical trials and it demonstrated ideal pharmacokinetics and metabolism.^28,29

Our group successfully automated the first MT-tracking PET radioligand, [¹¹C]MPC-6827, that crosses the blood–brain barrier and conducted evaluations of in vitro, in vivo, and ex vivo imaging utility in rodent models of AD.³⁰ Our mechanistic experiments demonstrated that [¹¹C]MPC-6827 selectively binds to destabilized MTs. We showed the high brain penetration, favorable pharmacokinetics, and excellent test–retest properties of [¹¹C]MPC-6827 in normal healthy nonhuman primates (NHPs).^31,32 On the basis of these encouraging preliminary studies of [¹¹C]MPC-6827 in rodents³³ and NHPs,^31,32 we went on to evaluate its imaging potential in an established NHP model of AD.

The genetic composition and phylogenetic assembly of nonhuman primates (NHPs) are similar to humans. They share anatomical and physiological characteristics and molecular pathways with humans. Vervets (Chlorocebus sabaeus) are a class of NHP that has similar genome sequences and develop several age-related disorders that are similar to those seen in humans.³⁴ Therefore, vervets are considered as a well-established translational model for early AD-like neuropathology and symptomatology, as well as other neurodegenerative diseases.^35–39 We and other laboratories have investigated the long-term effects of age on blood and cerebrospinal fluid (CSF)-based biomarkers^34,40 and have tested novel drug targets, biomarkers, and their associated pharmacological effects^41–44 in vervets. Vervets are, thus, an exceptional translational model for understanding AD pathophysiology. Here, we report the preliminary proof-ofconcept results of PET imaging with [¹¹C]MPC-6827 in four females (20–29 years) with high or low CSF Aβ₄₂ levels, which correlate with Aβ plaque burden in this species, similar to humans.⁴⁵

2. MATERIALS AND METHODS

2.1. Radiochemistry.

[¹¹C]MPC-6827 radiochemistry followed our reported procedures.^30,54 Briefly, it was produced in a GE FXM automated radiochemistry module by reacting the corresponding precursor desmethyl MPC-6827 with [¹¹C]MeI (produced in a GE FxMeI module) bubbled in a NaOH/DMF solvent system at room temperature and heated at 80 °C for 5 min. The compound was then purified by standard [acetonitrile (ACN)–ammonium formate buffer] semiprep HPLC and C18 SepPak (10% ethanol–saline) elution methods. Chemical and radiochemical purity of [¹¹C]MPC-6827 was determined by quality control (QC)-HPLC before animal PET imaging and autoradiography studies proceeded.

2.2. Nonhuman Primates PET/CT Imaging.

All animal housing and handling and all experimental procedures were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (2011) and approved by the Wake Forest University School of Medicine Animal Care and Use Committee (ACUC). Environmental enrichment of all animals followed ACUC’s Non-Human-Primate Environmental Enrichment Plan.

PET/CT imaging of [¹¹C]MPC-6827 was performed in four female adult vervets (ages 20.1, 26.5, 28.1, and 29.3 and weight 6.5–9.5 kg). All vervets were fasted overnight prior to the day of scanning, anesthetized using ketamine (10 mg/kg, im), and transported to the PET scanner suite. Each monkey was administered isoflurane (3–5%) via nose cone until intubated with an endotracheal tube. Throughout PET scanning, 1.5% isoflurane in oxygen was maintained. [¹¹C]MPC-6827 was injected using a catheter inserted into an external saphenous vein. All vervets were continuously monitored for vital signs, such as heart rate, blood pressure, respiration rate, and temperature (maintained at 40 °C with a water-circulating heating pad), throughout the scanning procedure.

First, a low-dose CT-based attenuation correction scan was performed. Next, all the vervets were subjected to dynamic brain PET scanning (0–120 min) with an intravenous dose of 0.37 ± 0.03 GBq of [¹¹C]MPC-6827 using a 64 slice GE Discovery PET/CT scanner.⁵⁵ Image reconstruction of the acquired emission data for each frame included full quantitative corrections, including attenuation, and were reconstructed into 2 × 30 s, 3 × 1 min, 5 × 2 min, 4 × 4 min, and 9 × 10 min frames.^56,57

We focused our image analysis on those commonly studied brain regions most affected by aging and AD-related pathology in NHPs^58,59 and humans.⁶⁰ To identify these regions, magnetic resonance imaging (MRI) was used to acquire anatomical images of all four monkeys. Anesthesia was maintained with ketamine (15 mg/kg im) during the scanning procedure, and 3D MRI brain images were acquired with a 3T Siemens Skyra MRI scanner using typical NHP brain parameters (TE 3.32, TR 2700, flip angle 8, bandwidth 190 Hz/Px, FOV 128 mm, 256 × 256 matrix, slice thickness 0.5 mm).^55,56 T1-weighted whole brain images were used to anatomically define spherical regions of interest (ROIs). PET images were coregistered with MRI, and fused PET/MRI data were analyzed using PMOD Biomedical Image Quantification Software (version 3.5; PMOD Technologies).^27,61,62 Standardized uptake values (SUVs) of the whole brain and selected regions (cerebellum, frontal lobe, hippocampus, occipital lobe, parietal lobe, temporal lobe, and thalamus) were calculated by dividing the tracer concentration in each pixel by the injected dose per body mass. Time–activity curves (TACs) were generated from SUV values using the PMOD NEURO (Ver 3.5, PMOD Technologies LLC, Zurich, Switzerland) software analysis tool.⁶²

2.3. CSF Sampling and Measurement of CSF Aβ₄₂.

CSF samples were taken ~25–30 days prior to PET studies by inserting a 22-gauge needle percutaneously into the cisternal space while the ketamine-sedated animal was maintained in a lateral recumbent position, as previously described.⁶³ Approximately 1–1.5 cm³ of spinal fluid was obtained and frozen at −70 °C until metabolite determinations were made. Aβ_1–42 were measured in first-thawed CSF using a Luminex-based INNO-BIA Alzbio3 assay. Coefficients of variation of high and low kit controls for Aβ₄₂ were <10%, and the mean coefficients of variation across all samples was 11%–12% for Aβ_1–42.

2.4. Autoradiography Studies.

In vitro autoradiography studies were carried out on post-mortem brain sections of all four vervets following reported protocols.⁶⁴ Briefly, sections of selected regions of the brain (prefrontal cortex, frontal cortex, midbrain/thalamus, pons, hippocampus, and cerebellum) were mounted on glass slides (Superfrost Plus slides, Fisher Scientific, Waltham, MA) and air-dried for 30 min. The slides were then incubated in PBS (pH 7.4) for 10 min to remove any endogenous binding. For blocking experiments, nonradioactive MPC-6827 (200 μM) was added 30 min prior to radiotracer treatment. All the slides were air-dried and dipped in a solution containing ~0.5 MBq of [¹¹C]MPC-6827 in PBS and incubated for 30 min. The slides were then washed with PBS (4×) and water (1×) and quickly air-dried. The radiotracer-treated brain tissue slides were exposed to a radioluminographic imaging cassette BAS-IP SR 2025 (GE Healthcare) for 12 h at −20 °C and scanned with a GE Amersham Typhoon scanner (25 μm pixel size). Autoradiographs were analyzed using an ImageQuant TL 8.2. ROIs were manually drawn, and specific binding was calculated and expressed as photostimulated luminescence signals per square millimeter (PSL/mm²) using MCID core 7.1 software.

3. RESULTS AND DISCUSSION

Whole-brain SUVs of [¹¹C]MPC-6827 were plotted to study the correlation of MT dysregulation with CSF Aβ₄₂ and age (Figure 1). A negative correlation was observed between [¹¹C]MPC-6827 SUV and CSF Aβ₄₂ (Figure 1A), and a positive correlation was observed between [¹¹C]MPC-6827 SUV and age (Figure 1B). Vervets with low CSF Aβ₄₂ levels (representing higher plaque burden) showed greater radiotracer brain uptake than those with relatively high CSF Aβ₄₂ values (presumed low plaque burden) (Figure 1C). The TAC patterns in both younger and older vervets were similar: uptake peaked within the first 5–10 min of radiotracer injection and gradually washed out by 120 min (Figure 2).

Figure 1.

Correlation of [¹¹C]MPC-6827 with (A) CSF Aβ₄₂ and (B) age. (C) Representative PET/CT brain images of control (#1450) and AD-like (#1320) NHPs following an iv injection of [¹¹C]MPC-6827 (0.37 ± 0.03 GBq) for 0–120 min.

Figure 2.

Representative control NHP (n = 2) and AD-like NHP (n = 2) whole brain–time activity curves (TACs) from dynamic 0–120 min PET images from monkeys injected with 0.37 ± 0.03 GBq of [¹¹C]MPC-6827.

Analysis of SUVs from PET imaging data provided a clear understanding of differences in the uptake of the individual region. In the whole brain, the uptake was 13% more in NHP models of AD monkeys than that in cognitively normal monkeys. In the individual regions, 2.34%, 14.7%, and 6.4% more uptake was observed in frontal lobe, hippocampus, and occipital lobe, respectively, in NHP models of AD monkeys than cognitively normal monkeys (Figure 3), whereas, there was no change in the temporal lobe, and slightly less uptake was observed in thalamus, cerebellum, and parietal lobe regions.

Figure 3.

Representative standard uptake values (SUVs) from different regions of brain obtained from dynamic 0–120 min PET scans from control NHP (n = 2) and AD-like NHP (n = 2).

To confirm these PET imaging results, brain tissues from the same four vervets were used for in vitro autoradiography experiments. Vervet brains with low CSF Aβ₄₂ had greater [¹¹C]MPC-6827 uptake (Figure 4A,B) in the prefrontal cortex, pons, and hippocampus regions than those with high CSF Aβ₄₂ values, whereas the midbrain/thalamus, frontal cortex, and cerebellum exhibited slightly low uptake compared with cognitively normal monkey brains (Figure 4B).

Figure 4.

In vitro autoradiograms of different regions of the brain of AD-like NHP (#1320) (A) and its regional quantification (B) of [¹¹C]MPC-6827 from control and AD-like NHPs (n = 2/group). *p < 0.033; **p < 0.002; ***p < 0.001.

Age and AD-related neuropathological changes typically appear well before the onset of symptoms in what are often called preclinical or prodromal phases of disease.⁴⁶ As the preclinical period is often long, translational models are needed to identify early signatures of the pathological decline. The use of NHP models of late-life brain change, such as those evaluated here, enables the ability to not only detect such early changes but also to develop novel biomarkers that can accurately identify and predict such changes as early in the disease course as possible. NHP animal models and methods provide the critical ability to assess in vivo biomarkers alongside valuable ex vivo autopsy and autoradiographic data, thereby providing mechanistic understanding and neuropathological validation of imaging findings.⁴⁷

It is important to investigate the link between amyloid deposition—a critical initiating step in the neuropathological cascade of AD⁴⁸ and MT destabilization. Amyloid pathology has been shown to impact MT stability through multiple taudependent mechanisms.^49,50 In this study, we used CSF-based levels of Aβ₄₂ as an index of brain amyloid plaque burden and demonstrated the association between amyloid burden and MT destabilization assessed in vivo using [¹¹C]MPC-6827 PET for the first time.

Our prior mechanistic studies in tau knockout mouse models⁵¹ showed that [¹¹C]MPC-6827 is more selective for destabilized MTs, and this preliminary investigation in four vervets with different CSF Aβ values also supports this interpretation. High-radiotracer uptake was observed in regions commonly associated with high tau and amyloid plaque densities.^52,53 Current preliminary results encourage the study PET imaging of the [¹¹C]MPC-6827 radiotracer in more subjects of the AD model of NHPs to validate the results statistically and to establish the MT destabilization model for the AD pathogenesis. Also, future studies including longitudinal study of the AD model of NHPs with the [¹¹C]MPC-6827 radiotracer may further strengthen the present concept. Overall, our preliminary results validate MT destabilization as an important biomarker to study AD pathogenesis.

4. CONCLUSION

In the current study, we investigated the use of [¹¹C]MPC-6827, a novel PET imaging radiotracer for MT destabilization, in a NHP model of AD. We found that [¹¹C]MPC-6827 uptake was inversely related to CSF Aβ₄₂ levels and directly related to age. These preliminary results suggest that MT destabilization increases with AD burden (i.e., high plaque and low CSF Aβ₄₂ values) and that MT PET may provide a useful biomarker for NHP and human studies of late-life neurodegenerative pathology, such as AD.