Quantification of [ 11 C]ABP688 binding in human brain using cerebellum as reference region: biological interpretation and limitations

Michele Stanislaw Milella, Luciano Minuzzi, Chawki Benkelfat, Jean‐Paul Soucy, Alexandre Kirlow, Esther Schirrmacher, Mark Angle, Jeroen Verhaeghe, Gassan Massarweh, Andrew J. Reader, Antonio Aliaga, José Eduardo Peixoto-Santos, Marie‐Christine Guiot, Eliane Kobayashi, Pedro Rosa‐Neto, Marco LeytonmedRxiv 2024 · DOI

TL;DR

Researchers have validated the cerebellum as a reliable "reference region" for quantifying mGluR5 receptors in the human brain using $[^{11}C]ABP688$ PET imaging. Despite the physical presence of the receptor protein in the cerebellum, the specific allosteric site targeted by the tracer is largely unavailable, allowing for accurate, non-invasive brain imaging without the need for painful arterial blood sampling.

What Problem Does This Paper Address?

To measure the density of neuroreceptors in a living brain using Positron Emission Tomography (PET), scientists usually need a "gold standard" measurement of the tracer's concentration in the blood over time. This requires an invasive arterial line—a procedure that is uncomfortable for patients and resource-heavy for clinics.

An alternative is the "reference tissue method," which compares a target region (like the hippocampus) to a reference region that is ideally devoid of the receptors being studied. For the metabotropic glutamate receptor type 5 (mGluR5), the cerebellum is the traditional candidate for a reference region. However, previous primate studies suggested the cerebellum might actually contain significant mGluR5 levels, which would "contaminate" the reference signal and lead to underestimations of receptor density elsewhere. This paper seeks to settle the debate: Is the human cerebellum a valid reference for $[^{11}C]ABP688$ PET scans?

Methods

The authors employed a dual-pronged approach, combining postmortem tissue analysis with live human imaging:

  1. In Vitro (Postmortem): They examined human brain slices using two techniques:
  2. Quantitative Autoradiography: Using a radioactive version of the tracer ($[^3H]ABP688$) to see exactly where it "sticks."
    • Immunohistochemistry (IHC): Using antibodies to detect the physical presence of the mGluR5 protein itself.
  3. In Vivo (Live PET): Six healthy volunteers underwent PET scans using a High-Resolution Research Tomograph (HRRT).
  4. The "Gold Standard": They performed full arterial blood sampling to create a metabolite-corrected arterial input function (AIF).
  5. The Comparison: They compared the AIF-derived data against non-invasive models (SRTM and NIGA) that used the cerebellum as the reference region.

Datasets

  • Postmortem Tissue: Brain specimens from the Douglas Mental Health University Institute Brain Bank ($n=5$ for autoradiography; $n=6$ for IHC).
  • Live Human Subjects: Six healthy volunteers (2 women, 4 men; mean age 24.5) recruited via community advertisements.
  • Benchmarks: The study compared the 2-tissue compartment model (2TCM) using arterial data against the Simplified Reference Tissue Model (SRTM) and Non-invasive Logan Graphical Analysis (NIGA).

Key Insights and Ideas

  • The "Presence vs. Availability" Dissociation: This is the study's most striking finding. IHC showed that the cerebellum has only 26% less mGluR5 protein than the hippocampus. However, autoradiography showed it has 94% fewer binding sites for the tracer.
  • Conformation Matters: The authors suggest that while the receptor protein exists in the cerebellum, it likely exists in a different "shape" (conformation) or isoform that hides the specific pocket where $[^{11}C]ABP688$ binds.
  • Validation of Non-Invasive PET: Because the tracer doesn't "stick" to the cerebellar receptors, the cerebellum acts as a nearly perfect "blank" for the mathematical models. This confirms that researchers can skip the arterial line for mGluR5 studies in humans without losing significant accuracy.

Key Results

  • High Correlation: Regional binding potential ($BP_{ND}$) values calculated using the cerebellum as a reference were highly correlated ($R^2 > 0.9$) with those derived from the invasive arterial method.
  • Low Bias: The non-invasive Simplified Reference Tissue Model (SRTM) showed an average bias of only about 7% compared to the invasive gold standard.
  • Regional Specificity: The highest receptor distribution volumes ($V_T$) were found in the prefrontal cortex and anterior cingulate cortex, while the lowest were in the cerebellum, as expected for a valid reference region.

Limitations and Future Work

  • Small Sample Size: With only six live subjects, the study is a robust "proof of concept" but would benefit from larger cohorts.
  • Healthy vs. Clinical Populations: The study was conducted on healthy volunteers. In certain diseases (e.g., cerebellar atrophy or specific neurodegenerative disorders), the cerebellum's status as a valid reference might change due to structural or biochemical shifts.
  • Tracer Isomers: The tracer used was a 1:8 ratio of cis/trans isomers. While the authors argue this doesn't affect the correlation between methods, using a pure "active" isomer might change the absolute values of the measurements.
  • Future Direction: The authors suggest that future studies should investigate why the cerebellar mGluR5 receptors are "invisible" to allosteric tracers—potentially uncovering new insights into receptor isoforms and their roles in different brain regions.

Summary updated

Abstract

Abstract In vitro data from primates provide conflicting evidence about the cerebellum’s suitability as a reference region for quantifying type 5 metabotropic glutamate receptor (mGluR5) binding parameters with positron emission tomography (PET). To address this, we first measured mGluR5 density in postmortem human cerebellum using [ 3 H]ABP688 autoradiography (n=5) and immunohistochemistry (n=6). Next, in vivo experiments were conducted in healthy volunteers (n=6) using a high-resolution PET scanner (HRRT) to compare [ 11 C]ABP688 binding potential (BP ND ) values obtained with reference tissue methods and the two-tissue compartment model vs. metabolite-corrected arterial input function. The postmortem data showed that, relative to the hippocampus, the cerebellum had 26% less mGluR5 immunoreactivity and 94% fewer [ 3 H]ABP688 binding sites. In vivo brain regional [ 11 C]ABP688 BP ND values using the cerebellum as a reference region were highly correlated with BP ND values and distribution volumes derived by arterial input methods (R 2 > 0.9). The absence of cerebellar allosteric binding sites might reflect the presence of distinct mGluR5 isoforms or conformational state. Together with our PET data, these results support the proposition that [ 11 C]ABP688 BP ND using cerebellum as a reference region provides accurate quantification of mGluR5 allosteric binding in vivo .
BibTeX
@misc{milella2024quantification,
  author = {Michele Stanislaw Milella and Luciano Minuzzi and Chawki Benkelfat and Jean‐Paul Soucy and Alexandre Kirlow and Esther Schirrmacher and Mark Angle and Jeroen Verhaeghe and Gassan Massarweh and Andrew J. Reader and Antonio Aliaga and José Eduardo Peixoto-Santos and Marie‐Christine Guiot and Eliane Kobayashi and Pedro Rosa‐Neto and Marco Leyton},
  title = {{Quantification of [ <sup>11</sup> C]ABP688 binding in human brain using cerebellum as reference region: biological interpretation and limitations}},
  year = {2024},
  doi = {10.1101/2024.02.12.24302279},
  howpublished = {medRxiv},
}
Figure