For several years, we have focused on developing much-needed effective, sensitive, and reliable radioligands capable of imaging neuroinflammation. These efforts included the development of now widely-used radioligands for translocator protein (TSPO), as described in past annual reports. Building on this work, we directed our efforts towards the cyclooxygenase (COX) system, which is implicated in the pathophysiology of brain diseases, including Alzheimers Disease and major depressive disorder, and is another potential biomarker for neuroinflammation. The COX system comprises two isoforms, COX-1 and COX-2, which are key enzymes in neuroinflammation. To study these COX isozymes in vivo, we recently developed two novel PET radioligands: 11C-PS13, which demonstrated potent selectivity for COX-1 (IC50=1 nM) compared to COX-2 (IC50>1,000 nM) and 11C-MC1, which was potent and selective for COX-2 (IC50=1 nM) compared to COX-1 (IC50>1,000 nM). We conducted two studies to fully assess the utility of both ligands. The first sought to image the in vivo distribution of COX-1 in brain and peripheral organs using 11CPS13 in humans and rhesus monkeys as well as evaluate the in vivo selectivity of 11CPS13 for COX-1. Building on the first, the second study examined whether 11C-MC1 could image COX-2 in a model of neuroinflammation using intracerebral lipopolysaccharide (LPS) injection in monkeys. In the COX-1 study, approximately 740 and 185 MBq of 11CPS13 were injected intravenously into rhesus monkeys, followed by dynamic PET scans. Two brain and 11 whole-body scans were obtained in human subjects, and 16 whole-body scans were obtained in monkeys. To measure the in vivo selectivity of 11CPS13, scans were also performed in monkeys pretreated with non-radioactive drugs preferential for COX-1 or COX-2; concurrent blood samples were obtained during the PET scans. In the COX-2 study, LPS was injected into the right putamen of four monkeys to elicit an inflammatory response. In two monkeys, LPS injection was repeated after a month. Dynamic brain PET scans were acquired for two hours both pre- and post-LPS injection (on Day 1). Approximately 220 MBq of radioligand was injected intravenously into the monkey before each scan. Blocking studies were also conducted with non-radioactive PS13 or MC1 (0.3-1 mg/kg) to confirm the specific uptake of the radioligands in the brain. Full quantitation with arterial sampling was done to measure distribution volume (VT). To confirm in vivo PET results, postmortem monkey brain tissues were obtained after PET scans and fluorescent in situ hybridization, immunostaining, and Western blot analyses were done. We found that 11CPS13 showed specific uptake in most major organs including spleen, gastrointestinal tract, kidney, and brain under baseline conditions in both humans and monkeys. In monkeys, 11CPS13 uptake in these organs was blocked by preferential inhibitors for COX-1, but not COX-2. In human brain scans, 11CPS13 showed prominent uptake in the hippocampus, occipital cortex, thalamus, and brainstem. In the COX-2 study that used an LPS model of neuroinflammation, a widespread increase in 11C-MC1 uptake was observed in brain after the first LPS injection, which was displaced by non-radioactive MC1. An even more remarkable increase in specific uptake of 11C-MC1 was noted following the second LPS injection, especially at the injection site. In postmortem brain after the first injection, in situ hybridization showed upregulation of COX-2 transcript, which was localized primarily in neurons. Immunostaining also showed upregulation of COX-2 protein. In postmortem brain after the second injection, a necrotic lesion with white blood cells was observed, comprising primarily macrophages and neutrophils expressing COX-2. Taken together, our results indicate that COX-1 was constitutively expressed in major organs; the in vivo selectivity of 11CPS13 was also well demonstrated by pharmacological blockade in monkey. 11C-MC1 was able to image COX-2 upregulation in monkey brain. Notably, 11CPS13 and 11C-MC1 are the first radioligands for COX-1 and COX-2, respectively, that act directly at these targets. In tandem, these ligands could be used to measure COX-1 in healthy conditions and COX-2 in inflammatory disorders, and could also be used to assess drug delivery and in vivo selectivity of NSAIDs in therapeutic trials.
Showing the most recent 10 out of 101 publications