Imaging Neuronal Pathways of the Rat Basal Ganglia System with 52Mn PET and MEMRI

Abstract

Introduction: The basal ganglia are components of neuronal circuits responsible for important functions, such as learning, memory, motor control and motivation. Disturbances of these functions are symptoms of neuropsychological diseases, including the Parkinson’s disease, attention deficit hyperactivity disorder, depression and addiction. Since a changed neuronal activity in the involved pathways has been observed in all these diseases, it is important to develop a sensitive tool for functional studies of the neuronal tracts in vivo. The present work proposes 52Mn for tracing neuronal pathways of the rat brain in an activity-dependent manner with positron emission tomography (PET). Several crucial aspects of this application, including the dosage, image quality and quantification, as well as pharmacological properties of the tracer, have been carefully evaluated. Methods: Firstly, the relationship between the dose of non-radioactive Mn2+ injected into the ventral tegmental area (VTA), the signal enhancement it provides in magnetic resonance (MR) images and the dopaminergic toxicity, was assessed. Secondly, a 52Mn phantom study was conducted to evaluate the quality of the obtained PET images. Next, a behavioral test and immunohistochemical staining techniques were used to evaluate the potentially harmful impact of the radioactivity doses in the range from 20 to 170 kBq on motor control system, dopaminergic neurons and the DNA damage. Having defined the non-toxic dose, [11C]methylphenidate and [11C]flumazenil PET measurements were performed to assess the effect of a single intracerebral injection of Mn2+ and 52Mn on the dopaminergic and GABAergic systems, respectively. Finally, the activity-dependence of the 52Mn neuronal transport was studied with pharmacological agents which modulate the neuronal activity. Additionally, an experimental setup for PET with an online measurement of the radioactivity blood level with a blood sampler (BS) was established. It was subsequently used to investigate the impact of different techniques of obtaining the arterial input function (AIF) on the kinetic parameters (KPs) of [18F]fluorodeoxyglucose ([18F]FDG). The KPs estimated using the AIF derived from the manually collected and dispersion-corrected blood samples served as the reference approach. Results: 45 nmol of Mn2+ provided a visible, although not statistically significant, change in the T1 maps 24 h after administration. However, it also caused a dopaminergic lesion at the injection site. The 0.5 nmol dose did not lead to the toxic effect, but it was not sufficient for MR imaging. Furthermore, 50 nmol of Mn2+ significantly reduced binding potential of [11C]flumazenil in the frontal cortex ipsilateral to the injected striatum at 2 days, but not at 4 weeks post-injection. There was no impact of the metal on the binding potential of [11C]methylphenidate. 24 h after the injection of 170 kBq of 52Mn into the VTA, the tracer was clearly visible along the mesolimbic and mesostriatal pathways in the PET images. The quantitative analysis confirmed its accumulation also in other expected brain areas. However, the dose led to an impaired rotation behavior and a dopaminergic lesion 4 weeks post-injection. These side effects were avoided by reducing the dose to 20 kBq. Moreover, 150 kBq of 52Mn, but not the 30 kBq dose, induced a substantial DNA damage. The neuronal transport of 52Mn was not affected by any of the tested pharmacological agents, except for the non-radioactive Mn2+ which substantially reduced 52Mn concentration in the analyzed brain regions. However, different transport rates and distribution patterns of the tracer were observed depending on the injected subnucleus of the VTA. In the BS study, the sensitivity of the device was 23 %. Using a non-corrected image-derived AIF resulted in large differences of the KPs from the reference values. The deviations of the KPs estimated with the BS-based approach were 3-5 % if 10 min of the BS recording was combined with one manually collected blood sample to obtain the AIF. However, using the plasma/(whole blood) activity ratio led to an 8 % overestimation of the k3. Conclusions: Imaging neuronal tracts of the rat brain with 52Mn PET is possible. In order to avoid harmful effects of the applied radioactivity, doses in the range of 20-30 kBq should be used. The activity-dependence of the 52Mn neuronal transport could not be confirmed and should be farther investigated. The online recording of the radioactivity level with the BS assures an accurate measurement of the peak activity after a bolus injection and prevents a blood loss. The obtained AIF allows reliable kinetic modeling if one manual blood sample is included. However, the differences in the estimated KPs lower than 10 % should be interpreted carefully provided the plasma/(whole blood) ratio was used.