Studies of neuroinflammation using PET in an animal model of schizophrenia

Introduction: Nowadays, the etiology of schizophrenia remains unknown. However, it has recently been pointed out that neurodevelopmental alterations during gestation could be involved in the manifestation of this pathology at adulthood. Thus, a disruption in the cytokine balance of developmental brains could trigger the activation of microglial cells in the central nervous system (CNS) and initiate an inflammatory response. The work has a two-fold aim: first to validate a correct methodology to detect the presence of uptake differences in [¹¹C]-PK11195-PET images; and second, to study the neuroinflammatory changes in an animal model of maternal immune stimulation (MIS) of schizophrenia by means of this imaging technique.

Methods: In gestational day 15, Poly I:C (Poly) or saline (Sal) were injected to pregnant Wistar rats. Male offspring (12 saline and 10 Poly) were imaged in a preclinical PET/CT scanner at two temporal points (adolescence and adulthood). [11C]-PK11195 was administered and brain dynamic images were acquired. Differences in uptake between groups were assessed with Statistical Parametric Mapping software (SPM12).

Results: Static images from frames 21-26 showed greater t-values compared with frames 26-34. Moreover, the size of a Gaussian smoothing kernel influenced in the size of activation area and t-value, increasing the area but disminishing the t-value at higher smooth (3 vs 2.5). Images normalized to a reference region showed more activated regions, higher t-values and lower p-values compared to SUV normalization. At adolescence, Poly I:C offspring showed reduced uptake in the septum and the substantia nigra compared to saline offspring. At adulthood, Poly I:C offspring showed increased radiotracer uptake in the brainstem, thalamus and cortex; and reduced uptake in the cerebellum and septum.

Conclusion: The appropriate methodology for our images consisted in the used of: static images from frames 21-26, Gaussian filter of 2.5 times the voxel size and reference region normalization. Moreover, exposure to maternal viral infection exerts a negative impact on brain maturation that is associated with increased glial activity in the brainstem, precisely in the ventral tegmental area and raphe nucleus.