Challenges in the differentiation of midbrain raphe nuclei in neuroimaging research

Abstract

The work by Takahashi et al. (1) recently reported a negative association between brainstem serotonin transporters and personality traits such as honesty and rejection to unfairness. They interpret their locus within brainstem to include the dorsal raphe nucleus (DRN) (1). Other than DRN, the median raphe nucleus (MRN) also projects to the forebrain and may be equally important in the serotonergic modulation of cortical activity. Hence, several studies have focused on these two midbrain raphe nuclei with respect to receptor and transporter distributions or anatomical connections. Small nuclei such as DRN and MRN are, however, exceptionally difficult to localize in human imaging research in vivo, such as positron emission tomography (PET).The works of Baker et al. (2, 3) provided insight into the distribution and location of serotonergic neurons within MRN and DRN in the postmortem human brain ( Fig. 1 A and B). Rostrocaudally, the DRN extends at the level of accessory oculomotor nucleus to the motor trigeminal nucleus. At the level of the trochlear nucleus ∼7-5 mm above the isthmus, the DRN exhibits its fountain-shaped appearance in a transverse section with a ventrodorsal and mediolateral extension of about 5 mm. The DRN above the isthmus includes the majority of neurons, with its subparts containing the greatest number of serotonergic neurons. At the level of the isthmus, the DRN narrows extensively to just 0.5 mm in the mediolateral extension and nestles up against the fourth ventricle. Farther caudally, the DRN divides into two staves and extends 14 mm below the level of the isthmus. For reliably delineating the DRN, a circular region of interest (ROI) may be placed at the level between superior and inferior colliculus and not extending 6 mm in diameter in a transverse plain in two slices. The ROI then may include up to 14 voxels given a voxel size of 2 × 2 × 2 mm of a PET image in MNI (Montreal Neurological Institute) standard space (Fig. 1C). In comparison, the MRN in the pontine tegmentum extends rostrocaudally from the level of the decussation of the superior cerebellar peduncle (1 mm above isthmus) down to the motor trigeminal nucleus ∼18 mm below isthmus. Most serotonergic neurons of the MRN lie within the first 10 mm below the level of the isthmus. In a transverse plain, the MRN appears with a ventrodorsal extension of about 6 mm and a mediolateral extension of about 2 mm. A reliable delineation of MRN may, therefore, provide a ROI in two slices within the rostral one-third of the pons, not extending 8 voxels.Given figure 2a in the work by Takahashi et al. (1), we propose that they measured serotonin transporters within the MRN rather than the DRN and suggest that MRN may be easily mistaken in imaging studies as DRN. DRN and MRN differ in several aspects, including receptor distributions and afferent and efferent connections (4), which highlights the importance of differentiating and identifying them as two separate nuclei in neuroimaging research.

Monoaminergic neuropathology in Alzheimer’s disease

“…Therefore, expanding our understanding of the brainstem nuclei, particularly the LC and DRN involvement in the early stages of AD to functional concepts beyond neuropathological descriptions, will likely have a significant impact on diagnosis and tracking of AD progression as well as on development of biomarkers, novel therapeutic targets, and preventive strategies. One of the recent breaktroughs in that direction was the successful visualization of the DRN using PET imaging of 5-HT 1A receptor binding (Kranz et al, 2012). Other potential pharmacological interventions directed mainly at serotonergic system targets are expected to alleviate symptoms of neurodegeneration, and to expand our understanding of the relationship between monoaminergic systems and the pathogenesis of AD.…”

Functional connectivity of the dorsal and median raphe nuclei at rest

“…However, to perform seed-based FC, accurate in vivo segmentation of DR and MR are needed (Kalbitzer and Svarer, 2009). This presents a challenge (Kranz and Hahn, 2012), as the raphe nuclei are composed of sparse neurons surrounded by white matter and they have no well-defined boundaries visible in MRI (Baker et al, 1996, 1991a, 1990). …”

“…Based on the fact that the DR, excluding the caudal subnucleus, is about 57 mm 3 , we used a target volume of 115 mm 3 . The target volume used for the MR was 64 mm 3 , as suggested by Kranz & Hahn (2012). This procedure was applied to enforce local convergence; more lenient clustering methods, such as taking the maximum voxels within the search volume, led to structurally inhomogeneous ROIs, inconsistent with the morphology of the DR and MR. Once the iterative process completed, the seed was transferred onto the GD corrected structural MRI and an inverse gradient unwarping (i.e., reintroduction of gradient non-linearities) was applied to the ROIs to match gradient non-linearities present in the BOLD fMRI images.…”

Fluoxetine Administered to Juvenile Monkeys: Effects on the Serotonin Transporter and Behavior

“…Dynamic PET images were coregistered to averaged MRI templates from six monkeys in standardized space. Time-activity curves were generated using predefined regions of interest for both the neocortex and the hippocampus (27) and manually for the raphe (28). For the neocortex, five different cortical regions were combined and weighted for volume: the frontal, cingulate, temporal, parietal, and occipital cortices.…”