DPA-714 and most other microglial PET ligands bind to the outer mitochondrial membrane translocator protein 18 kiloDalton (TSPO, formerly known as the peripheral-type benzodiazepine receptor; Papadopoulos et al., 2006). Its function in the central nervous system is uncertain; it appears to be involved in a variety of cellular tasks including transport of cholesterol and other molecules. TSPO is not produced in the central nervous system under normal conditions, but rapidly appears in activated microglia. This makes it “by far the best biomarker for brain injury and inflammation for noninvasive imaging,” said Tomás Guilarte of Columbia University in New York. Neither Guilarte nor Wiley were involved in the April 25 study.

Spinal Spots
As reported in the Journal of Neuroscience paper, first author Galith Abourbeh imaged DPA-714 in the rat spinal cord. Doctors typically use magnetic resonance imaging to identify the spinal lesions typical of MS, but this technique misses out on the most subtle inflammatory pathology, Abourbeh said.

The researchers immunized rats with myelin basic protein to induce acute experimental autoimmune encephalitis, a common model for MS. Injecting DPA-714 labeled with radioactive fluorine-18, Abourbeh observed that immunized rats expressed fivefold more TSPO than control animals. “Using DPA-714, we could image and detect neuroinflammation in this model,” Abourbeh concluded. Guilarte commented that, while he would have liked to see the DPA-714 signal go down as the rats recovered their health—an analysis the authors did not include—the study is a “good first attempt.”

A Multitude of Markers
DPA-714 is one among more than a dozen potential TSPO tracers that have emerged in recent years. “There has been an explosion of groups trying to develop better ligands to image it,” Guilarte said. Researchers hope to improve upon the signal provided by the TSPO ligand PK11195, for decades the standard tracer. PK11195 has been applied in studies of Alzheimer’s (see ARF related news story on Cagnin et al., 2001), multiple sclerosis (Banati et al., 2000), amyotrophic lateral sclerosis (Turner et al., 2004), frontotemporal dementia (Cagnin et al., 2004), Parkinson’s (Gerhard et al., 2006), Huntington’s (Tai et al., 2007), as well as other neuroinflammatory conditions. However, PK11195’s characteristics make it less than ideal, Abourbeh said. The molecular makeup means that it requires carbon-11 as a radiolabel, with a half-life of only 20 minutes. In contrast, DPA-714 can be tagged with fluorine isotopes, with a more convenient half-life of nearly two hours.

In addition, PK11195 is not very specific for TSPO, Abourbeh said. Hence, scientists are testing a slew of other potential TSPO tracers, for example, CLINDE (Mattner et al., 2005), DAA1106 (Yasuno et al., 2008), SSR180575 (Chauveau et al., 2011), FEPPA (Wilson et al., 2008), CLINME (Boutin et al., 2007), vinpocetine (Vas et al., 2007), and many others (reviewed in Luus et al., 2009; Chauveau et al., 2008; James et al., 2006). DPA-714 has already been tried in people, including in a Bayer HealthCare trial attempting to differentiate people with probable AD from healthy participants. However, Bayer halted the study early because an interim assessment showed no difference between the two groups.

In other research, scientists attempted to correlate the PK11195 signal with that from Pittsburgh compound B (PIB; see ARF related news story on Kadir et al., 2011). In one such study, researchers reported that PK11195 signals correlated with the PIB label (Edison et al., 2008); in another, Wiley and colleagues discovered no such overlap (Wiley et al., 2009). PK11195 was probably not sensitive enough to pick up the amyloid-linked inflammation in his study, Wiley said. This variability in results is part of the problem with PK11195, said Agneta Nordberg of the Karolinska Institute in Stockholm, Sweden, who was not involved with the April 25 paper but has imaged astrocytes in early AD patients (see below).

Modern Challenges
“It would be so much better to get something more selective and more sensitive,” Wiley said. He thinks TSPO is not the way to go. One problem is that modern TSPO ligands do not work for all people. Researchers developing the TSPO ligand PBR28 have found a polymorphism in the human TSPO gene at position 147, normally an alanine but a threonine in the minor allele (Owen et al., 2012). People heterozygous at this locus exhibit low signals with PBR28 PET, and people homozygous for the threonine allele—approximately 10 percent of people (Fujita et al., 2008)—show no signal at all. The polymorphism appears to affect several modern TSPO ligands including PBR06, DAA1106, PBR111, and a DPA-714 analog (Owen et al., 2011). Researchers suspect that this discrepancy does not show up in PK11195 studies because that tracer is not highly specific for TSPO. Scientists are still trying to figure out how to properly analyze studies of people with different TSPO genotypes, Guilarte said.

Another problem with TSPO tracers is that they are not unique to microgliosis. At times, activated astrocytes turn it on as well (Ji et al., 2008). “I would rather spend time and effort to identify different molecules that bind to more specific targets of microglia and macrophages,” Wiley said. “Immunologists have identified tons of targets for microglia.” For example, he suggested, the marker CD68 would be a more logical tracer target than a mitochondrial protein of uncertain function and distribution. Potential non-TSPO targets include the cannabinoid type 2 receptor (Horti et al., 2010), cyclooxygenase-1 (Shukuri et al., 2011), and -2 (de Vries et al., 2008), CB2 and P2X7 (Yiangou et al., 2006), and metalloproteinases (Wagner et al., 2007). Guilarte and Wiley agreed that, while differentiating astrocytosis and microgliosis is important for scientific studies, medically it might not matter, since both indicate neuroinflammation.

Starry Studies
While much attention has focused on microgliosis, there is also a PET tracer for astrocytosis. L-deprenyl sticks to monoamine oxidase B, an enzyme on the outer mitochondrial membranes of astrocytes that metabolizes neurotransmitters (Fowler et al., 2005). Nordberg used deprenyl to discover that astrocytes are most strongly activated in people with mild cognitive impairment, even compared to subjects with full-blown Alzheimer’s (see ARF related news story on Carter et al., 2012). She has started a longitudinal study of people genetically at risk to develop AD, looking with both deprenyl and PIB to discover the earliest signs of pathology.

While some microglial tracers look promising, none quite fit the bill, and none are in regular clinical use, Wiley said. Once researchers have a good tracer, it could help them better understand the process of neuroinflammation, Nordberg said. She doubted markers for microgliosis or astrocytosis would be useful for diagnosis, because those processes are common to so many conditions. The greatest benefit, Wiley said, would be to use tracers to evaluate the efficacy of anti-inflammatory medications in clinical trials.—Amber Dance.

Reference:
Abourbeh G, Thézé B, Maroy R, Dubois A, Brulon V, Fontyn Y, Dollé F, Tavitian B, Boisgard R. Imaging microglial/macrophage activation in spinal cords of experimental autoimmune encephalomyelitis rats by positron emission tomography using the mitochondrial 18 kDa translocator protein radioligand [18F]DPA-714. J Neurosci. 2012 Apr 25;32(17):5728-36. Abstract