Background: Multiple sclerosis (MS) is a devastating disease of the central nervous system (CNS) that can only be diagnosed following the onset of severe clinical symptoms. Chronic neuroinflammation precedes the onset of clinical symptoms. The potential to diagnosis MS early has been demonstrated with PET tracers targeting 18kDa translocator protein (TSPO), where neuroinflammation was detected prior to the detection of lesions by MRI^1,2. TSPO is not an ideal biomarker of neuroinflammation because it lacks cellular specificity³. Cyclooxygenase-1 (COX-1) is reportedly localized to microglia in the brain³ and has been linked to disease progression in mouse models of MS^4,5. Herein we employ [¹¹C]PS13, a brain penetrating, highly potent and selective COX-1 radiopharmaceutical^6,7, to reveal changes in COX-1 density in the CNS of experimental autoimmune encephalomyelitis (EAE) mouse models, the standard animal model of MS fdsaa 

Methods: Radiosynthesis of [¹¹C]PS13 and dynamic PET imaging was performed as previously described, with minor modifications⁸. Wildtype (WT) C57BL/6 mice (2F/2M) and EAE mice with scores >3 (5F/3M) were injected with [¹¹C]PS13 (A_m = 50-100 GBq/µmol) via tail-vein catheter and dynamic PET scans were aquired for 60 min followed by acquisition of MRI for co-registration, PET scatter and attenuation corrections. Time-activity curves (TACs) were extracted from selected brain regions. Herein, we report data for the hippocampus and cervical spinal cord (SC). Binding potential (BP_ND) was estimated using Logan’s plot incorporating forebrain as a reference region, t*12 in Pmod V4.2. Immunofluorescence was performed to investigate the co-localization of COX-1 with Iba-1 and GFAP, biomarkers of activated microglia and astrocytes, respectively.

Results: TACs revealed higher [¹¹C]PS13 standard uptake values (SUV) in EAE compared to WT. The hippocampus had the highest SUV, and in  female mice (Fig. 1A and 1B) SUV was clearly higher in EAE compared to WT, though this was not apparent in the male groups (Fig. 1C and 1D) which showed no difference in SUV of the tracer in the hippocampus, though a slightly higher SUV was observed in the cervical spinal cord (SC) of male mice. This work reveals preliminary evidence of sex differences in COX-1 density in the brain. BP_ND estimates in the hippocampus (Fig. 1E) and cervical SC (Fig. 1F) revealed further evidence of sex differences where a trend is observed in WT female mice having higher BP_ND values than WT males, however, no difference in BP_ND was observed between male and female EAE mice. Immunofluorescence clearly shows regions in which COX-1 is extensively co-localized with Iba-1 compared to GFAP, indicating that COX-1 is localized in microglia (Fig. 1G).

Conclusions: Our preliminary data shows increased [¹¹C]PS13 uptake and BP_ND in EAE compared to WT mice, and sex differences. Future studies include further analysis of the PET data to determine volume of distribution. Immunofluorescence showing the co-localization of COX-1 with Iba-1 positive microglia, and not astrocytes, indicates the potential of [¹¹C]PS13 to specifically image microglia.

Innovation: A neuroinflammation PET imaging agent that is specific for microglia is being keenly sought. [¹¹C]PS13 represents a potentially microglial specific PET imaging agent.

Image/Figure:

Click to view full size

Image/Figure Caption:

Fig. 1. Time-activity curves (TACs) extracted from hippocampus (A) and cervical SC (B) of female WT and EAE mice. TACs extracted from hippocampus (C) and cervical SC (D) of male WT and EAE mice. BP_ND estimates in the hippocampus (E) and cervical SC (F) for WT and EAE mice. G) Immunofluorescent staining for COX-1 co-localization to Iba-1 positive microglia and GFAP positive astrocytes, showing extensive co-localization of COX-1 with Iba-1 compared to GFAP. 20X; scale bars denote 50 µm.

Author