Imaging the interaction of astrocytes and neurons

Comment

Immunological mechanisms and neuroinflammation play a key role in aging and neurodegenerative diseases. While astrocytes are essential for maintaining neuronal viability and brain function, they also play a key role in neuroinflammation which may contribute to neuronal loss. This has, for instance, been demonstrated with positron emission tomography (PET) and ¹¹C-deuterium-L-deprenyl (¹¹C-DED) in Alzheimer’s disease. ¹ On that background, Crine et al. ² hypothesized that the age-related decline of dopaminergic innervation, which is essential for cognitive functioning, is linked to astrocyte activation. They did not find that link but rather the opposite, an association of declining dopaminergic and astrocyte imaging, raising several questions.

Astrocyte imaging by ¹¹C-DED may not be the most sensitive indicator of neuroinflammation. Microglia is more closely associated with neuroinflammation, and an association between microglial activation and dopamine terminal loss has actually been demonstrated by PET in Parkinson’s disease. ³ Also, the enzyme monoamine oxidase B (MAO-B), which was used as an indicator of astrocyte activation by Crine et al., is involved in the catabolism of dopamine, though only secondary to MAO-A. Possibly, the observed positive correlation with dopaminergic innervation in elderly normal subjects could represent a sensible functional coupling of dopaminergic innervation and metabolism. Actually, a positive correlation between dopaminergic and serotonergic metabolism in brain has been reported. ⁴ As discussed by Crine et al., serotonergic neurons also express MAO-B, which means they could contribute to the observed activity measured by ¹¹C-DED PET. Still, the density of serotonergic neurons is much lower than that of astrocytes, supporting the use of ¹¹C-DED as an astrocyte marker.

Is it possible to get past these limitations? As an alternative to MAO-B, the imidazoline-2 binding sites (I₂BS) on astrocytes have successfully been targeted by the PET radioligand ¹¹C-BU99008, demonstrating a regional association with amyloid load in Alzheimer’s disease. ⁵ Using this tracer, increased astrogliosis related to aging was demonstrated. ⁶ There is few data using postmortem saturation radioligand binding, demonstrating regional differences of ¹¹C-BU99008 and ¹¹C-DED. ⁷ Thus, replicating Crine et al. using different tracers would be informative.

The large variety of astrocyte changes, activation and senescence, occurring with aging and neurodegeneration, ⁸ let alone their interaction with neurons, is by far too complex to be understood by single molecular imaging markers. Possibilities for further progress include multimodal imaging in combination with the analysis of fluid markers (in CSF or blood) of neuroinflammation and neurodegeneration. For instance, an association between plasma GFAP and regional ¹⁸F-SMBT-1, an alternative PET marker for MAO-B, and their dependency on brain Aβ load has been demonstrated. ⁹ While fluid markers cannot provide localization, they allow profiling using multiple markers, nicely complementing the spatial distribution of selected imaging markers.

Pooling data is an excellent tool to overcome the limits of individual imaging studies. Most widely known this route has been taken by the large Alzheimer’s Disease Neuroimaging Initiative, by the development of centiloid scaling for quantitative comparison of different beta amyloid tracers, and it has also been applied successfully to microglia tracers. ¹⁰ Accordingly, Crine et al. offer the availability of their data, for example, for multimodal studies and data pooling to investigate on a wider scale the interaction between glia and neuronal survival in aging and neurodegenerative diseases.