Peering into the brain to watch cholesterol metabolism in real time may have just become a reality, thanks to a new PET tracer. In the October 5 Science Translational Medicine, researchers led by Steven Liang at Emory University, Atlanta, described how 18F-Cholestify, which binds cytochrome P450 46A1, detected cholesterol breakdown in the mammalian brain. CYP46A1 converts cholesterol to 24-hydroxycholesterol, a form easily eliminated from the brain.

PET scans of healthy mice, monkeys, and young people found the highest tracer uptake in the cortex. In a mouse model of amyloidosis, more 18F-Cholestify bound to the hippocampus that it did in wild-type animals. “This PET tracer is a fascinating development that opens many possibilities for further research on cholesterol metabolism in neurodegenerative diseases,” Andreas Papassotiropoulos, University of Basel, Switzerland, told Alzforum.

  • 18F-Cholestify binds cytochrome P450 46A1, which converts cholesterol to 24-hydroxycholesterol.
  • In a mouse model of amyloidosis, tracer uptake was elevated in the brain.
  • PET signal was higher in young women than young men.

“Given the many uncertainties [of brain cholesterol metabolism] and the highly conserved nature of this metabolic pathway, this PET tracer will be a real game changer,” wrote David Russell, UT Southwestern Medical Center in Dallas (comment below).

One-fifth of all cholesterol in the human body exists in the brain. Four-fifths of that lies in the myelin sheaths that insulate neurons, and the rest in cell membranes (Zhang and Liu, 2015). However, since serum cholesterol cannot cross the blood-brain barrier into the brain, it must be made there, typically by astrocytes and neurons. What if they make too much? When that happens, neuronal CYP46A1 turns it into 24-hydroxycholesterol. This more water-soluble form easily slips through the cell membrane and across the BBB to exit the central nervous system. In people, about 6 mg of cholesterol a day is processed by this route. 

18F-Cholestify. This PET ligand binds CYP46A1 with high affinity. [Courtesy of Haider et al., Science Translational Medicine, 2022.]

If the amount of cholesterol present overwhelms CYP46A1, however, the lipid builds up. It forms deposits, typically as cholesterol esters, that encourage amyloidosis, tau aggregation, and neuronal death in mice and cultured cells (May 2022 news; Djelti et al., 2015; Sep 2001 news). In stem-cell-derived neurons from people with Alzheimer’s disease, boosting the hydroxylase activity of CYP46A1 with an allosteric modulator sent cholesterol droplets packing, reducing the amount of Aβ and phosphorylated tau, as well (Feb 2019 news). However, there has been no way to track this clearance mechanism directly in the brain—until now.

Co-first authors Ahmed Haider and Chunyu Zhao of Emory and Lu Wang at the First Affiliated Hospital of Jinan University, Guangzhou, China, made a small molecule that nuzzled tightly to CYP46A1’s active site, then added a fluorine-18 isotope to create the PET tracer 18F-Cholestify.

How does the tracer perform? Autoradiography of brain tissue slices from wild-type mice, rats, or rhesus monkeys treated with 18F-Cholestify showed that the tracer flocks to the cortex, hippocampus, striatum, and thalamus in rodents, and to the cortex, putamen, and caudate in primates. These areas were rich in CYP46A1 as measured by western blot of brain tissue. In contrast, the tracer barely bound to tissue slices from CYP46A1 knockout mice, or to tissue from mice treated with the CYP46A1 inhibitor soticlestat, indicating that the binding was specific for the enzyme.

In vivo, tracer uptake showed much the same binding. Within 10 minutes of intravenous injection, 18F-Cholestify lit up the cortex, hippocampus, striatum, and thalamus in mice, rats, and monkeys (see image below). In nonhuman primates, soticlestat dose-dependently dampened the PET signal. In mice, 18F-Cholestify uptake tightly correlated with 24-hydroxycholesterol concentration measured by mass spectrometry of brain tissue, but not with 24,25-epoxycholesterol, which is produced by other enzymes. All told, the authors concluded that the tracer measures CYP46A1-dependent cholesterol metabolism.

Rodent and Rhesus. On PET scans, 18F-Cholestify lights up CYP46A1 in the cortex (Cx), striatum (Str), thalamus (Th), and hippocampus (Hp) of a rat (left), mouse (middle), and monkey (right). Notably, hippocampal uptake is lower in primates than in rodents. The cerebellum (Cb) produces very little of the enzyme. [Courtesy of Haider et al., Science Translational Medicine, 2022.]

What about in people? In this study, four men and four women, age 22 to 31, had an 18F-Cholestify PET scan. The signal reached its peak intensity after just 15 minutes. Tracer uptake was highest in the cortex, thalamus, and basal ganglia, regions with the most CYP46A1 expression, according to western blots of postmortem tissue from healthy adults.

Notably, women had more 18F-Cholestify uptake in their putamen and caudate than did men, suggesting they muster higher baseline cholesterol clearance from the brain (see image below). However, the authors were cautious in drawing conclusions given the small sample size.

Sex Specific? Compared to a male participant (right), a female volunteer (left) had higher uptake of the 18F-Cholestify tracer in her putamen (black arrow) and caudate (white arrow). [Courtesy of Haider et al., Science Translational Medicine, 2022.]

Curious about what CYP46A1 might look like in AD, the scientists turned to 10- to 12-month-old 3xTg mice. These animals develop cortical amyloid plaques by six months and hippocampal neurofibrillary tangles between 12 and 15 months. Compared to wild-type animals, 3xTg mice intravenously injected with 18F-Cholestify had higher uptake in their hippocampi. The researchers think this might reflect the need to clear excess cholesterol and temper plaque accumulation and p-tau load.

Indeed, people who have mild cognitive impairment or early AD have more 24-hydroxycholesterol in their CSF than do healthy controls, and the CSF metabolite tracks with fluid p-tau181 levels (Papassotiropoulos et al., 2002; Leoni et al., 2006; Popp et al., 2013). “Before, you could only report CYP46A1 activity by measuring 24-hydroxycholesterol in the CSF; now, the PET ligand allows researchers to map it spatially in the brain,” said Mikael Simons of the Technical University Munich.

Simons thinks the high cholesterol turnover in the 3xTg mice could reflect neurodegeneration. Essentially, neuron death would release cell membrane fragments chock-full of cholesterol that must be cleared from the brain by conversion to 24-hydroxycholesterol. “It would be interesting to know if there is a correlation between levels of this molecule and neurofilament light, a marker of axonal degeneration,” said Simons. Papassotiropoulos agreed but noted that other interpretations remain open, too. “It might be a causal relationship. It might be an epiphenomenon. Or it might be specific to that particular transgenic model,” he said.

Papassotiropoulos added that the tracer could be used to characterize cholesterol clearance in the brains of people at different stages of AD. Scientists could then spot changes in different brain areas as the disease progresses. He thinks this tracer will help determine if CYP46A1 is hyperactive early in AD pathogenesis, as it is 3xTg mice, and how the enzyme’s activity relates to amyloidosis.

Rik van der Kant of Vrije Universiteit, Amsterdam, and Everard Vijverberg of the Amsterdam University Medical Center, noted that cholesterol does not regulate CYP46A1 levels, hence why the enzyme ticks up in 3xTg mice remains to be determined (comment below).  

Liang plans to partner with Emory’s AD Research Center to scan people with AD and healthy older adults using 18F-Cholestify PET; this work will start in early 2023.—Chelsea Weidman Burke

Comments

  1. I am familiar with the stellar research of Drs. Liang and Griffith, the senior authors of this study. Their development of a PET tracer for following brain cholesterol metabolism is significant, and, as they note, will allow many mysteries regarding this process to be solved. The work of Ingemar Björkhem in Sweden revealed that the conversion of cholesterol to 24-hydroxycholesterol was a route by which cholesterol could be metabolized in the brains of several mammals, including humans. My group isolated the enzyme that catalyzes this reaction (cholesterol 24-hydroxylase, gene symbol CYP46A1), and together with John Dietschy’s lab, we quantitated how much cholesterol is metabolized in the mouse and showed that disruption of the pathway by knocking out the gene caused learning difficulties in the species.

    Together this research revealed that metabolizing cholesterol in the adult mouse brain is important, but it is not clear why this is so or whether the same is true in humans. The CYP46A1 gene is very highly conserved in vertebrates (at least down to zebra fish), which suggests that it, and the pathway the encoded enzyme catalyzes, are likely important.

    There are several studies in the literature showing that 24-hydroxycholesterol levels in the blood (this product of the brain turnover pathway is secreted from the brain into the bloodstream and thereafter cleared by the liver) are altered in various diseases that affect the central nervous system, again hinting that cholesterol metabolism is important. There is little consensus, however, as to whether these levels are higher or lower than normal, or again, just what these data mean (for a review see Russell et al., 2009). 

    Given these many uncertainties in the field, and the highly conserved nature of the pathway, the PET tracer will be a real game changer. The number of noninvasive studies that can now be done is large, and (finally!) we will be able to learn why we metabolize about 6 mg of cholesterol every day through this route.

    References:

    . Cholesterol 24-hydroxylase: an enzyme of cholesterol turnover in the brain. Annu Rev Biochem. 2009;78:1017-40. PubMed.

  2. We are very excited by new developments that can help to monitor cholesterol metabolism in the living human brain. This study is highly impressive because it defines a novel and specific method to measure CYP46A1 levels in the human brain through PET imaging. Whether this novel 18F-CHL-2205 tracer can really be used to monitor changes in brain cholesterol homeostasis, as the title of the article suggests is, however, less clear.

    An increasing number of observations, including our own work in iPSC-derived neurons, have shown that excess levels of cholesterol in neurons can contribute to accumulation of both Aβ and p-tau (van der Kant et al., 2020). Cholesterol accumulates in the brains of AD transgenic mice, where a reduction of cholesterol levels has been shown to ameliorate pathogenesis. A few years back, the group of Irina Pikuleva discovered that low doses of the HIV drug efavirenz enhance cholesterol export from the brain by activating the CNS-specific protein CYP46A1. CYP46A1 converts cholesterol to 24-hydroxycholesterol, which can cross the blood-brain barrier into the blood.

    Based on these findings, and our own findings that CYP46A1 activation by efavirenz reduces p-tau levels in human iPSC-derived neurons and tau transgenic mice (unpublished), we are currently preparing a Phase 2a clinical trial in subjects with MCI or mild dementia due to AD. This dose-finding, randomized, double-blinded, multicenter study will evaluate different low doses of efavirenz and is set to start in spring of 2023. The goal is to define an optimal dose at which efavirenz activates CYP46A1. The primary outcome is target engagement (CYP46A1 activation), which will be indirectly evaluated by measuring the levels of 24-hydroxycholesterol in CSF and blood before, during, and after treatment.

    Methods that can more directly report on brain cholesterol levels, or brain cholesterol metabolism, in AD patients would therefore be of immense value for the field. In this well-crafted paper, the authors show that 18F-CHL-2205 can bind CYP46A1 in healthy volunteers. It is, however, unclear how this informs on cholesterol turnover or metabolism in humans. We have no doubt that information on CYP46A1 protein expression in brain gained by using 18F-CHL-2205 will be of high relevance for research into CYP46A1 biology. However, it is not clear whether 18F-CHL-2205 can be used to compare brain cholesterol levels between individuals (e.g., AD patients versus controls) and/or to monitor changes in brain cholesterol metabolism in response to interventions. This would really have been a game changer in the field.

    We are not planning to incorporate 18F-CHL-2205 PET scans into our trial because we are just finalizing the protocols and aim to start recruitment as soon as possible. 18F-CHL-2205 PET may be of interest for follow-up trials, yet we don’t directly see how information on CYP46A1 levels will help follow cholesterol metabolic changes in the brain. CYP46A1 is not regulated by cholesterol to our knowledge, and we would not expect it to change in the presence of CYP46A1 activators.

    18F-CHL-2205 PET might be helpful in possibly explaining differences among efavirenz responders and nonresponders, but we would assume everybody has some basal level of expression. In this respect it would have also been very interesting to see how CYP46A1 activators, such as efavirenz, alter the binding of 18F-CHL-2205.

    So, while this new tracer will certainly help to better understand CYP46A1 biology, and its relationship to AD pathogenesis, the search for a direct reporter of brain cholesterol levels in Alzheimer's patients continues.

    References:

    . Amyloid-β-independent regulators of tau pathology in Alzheimer disease. Nat Rev Neurosci. 2020 Jan;21(1):21-35. Epub 2019 Nov 28 PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Membrane Border Patrol: Cholesterol Stymies Tau Uptake, Aggregation
  2. Cellular Compartments Provide New Wrinkle—and a New Target?—in Cholesterol
  3. Cholesteryl Esters Hobble Proteasomes, Increase p-Tau

Research Models Citations

  1. 3xTg

Paper Citations

  1. . Cholesterol metabolism and homeostasis in the brain. Protein Cell. 2015 Apr;6(4):254-64. Epub 2015 Feb 15 PubMed.
  2. . CYP46A1 inhibition, brain cholesterol accumulation and neurodegeneration pave the way for Alzheimer's disease. Brain. 2015 Aug;138(Pt 8):2383-98. Epub 2015 Jul 2 PubMed.
  3. . 24S-hydroxycholesterol in cerebrospinal fluid is elevated in early stages of dementia. J Psychiatr Res. 2002 Jan-Feb;36(1):27-32. PubMed.
  4. . Are the CSF levels of 24S-hydroxycholesterol a sensitive biomarker for mild cognitive impairment?. Neurosci Lett. 2006 Apr 10-17;397(1-2):83-7. PubMed.
  5. . Cerebral and extracerebral cholesterol metabolism and CSF markers of Alzheimer's disease. Biochem Pharmacol. 2013 Jan 3; PubMed.

Further Reading

Primary Papers

  1. . Assessment of cholesterol homeostasis in the living human brain. Sci Transl Med. 2022 Oct 5;14(665):eadc9967. PubMed.