Introduction

Neuroinflammation is a key factor driving neurodegeneration [1]. During active neuroinflammation, resident brain macrophages termed microglia adopt a ‘reactive’ phenotype that generates a neurotoxic microenvironment in part due to the secretion of reactive oxygen species. We hypothesize that extracellular ROS released by reactive microglia during neuroinflammation will mediate the oxidation of Fe-PyC3A from the low-relaxivity Fe2+ state to the high-relaxivity Fe3+ state [2], generating strong MR signal enhancement. Here, we performed brain imaging with the oxidatively activated ROS-sensing MRI contrast agent, Fe-PyC3A, in a mouse model of lipopolysaccharide (LPS)-induced neuroinflammation for non-invasive detection of microglial changes.

Methods

A total of 24 C57BL/6 male mice and 4 gp91phox knockout (NOX-2 KO) mice (8-10 weeks old) received i.p. injection of either LPS (3 mg/kg) as a model of neuroinflammation or saline (3 mg/kg) as controls. After 24h, animals were anesthetized with isoflurane (1.5%) and imaged with a 4.7 Tesla Bruker MRI scanner equipped with a custom-built volume coil. Prior to i.v. injection of 0.3 mmol/kg Fe-PyC3A (N = 8 LPS, N = 6 saline) or 0.1 mmol/kg Gd-DOTA (N = 5 LPS, N = 5 saline), the blood-brain barrier was transiently disrupted via i.v. injection of 25% wt/wt mannitol. 2D multi-slice T1-weighted gradient echo images were acquired dynamically prior to and up to 20 minutes post-injection of Fe-PyC3A or Gd-DOTA. We selected hippocampus and cortex as regions of interest (ROIs) and analyzed the (post-pre) injection percentage change in signal intensity (SI%) from these ROIs and (post-pre) injection change in brain ROIs vs. muscle contrast to noise ratio (ΔCNR) from 2 -10 min after Fe-PyC3A or Gd-DOTA injection. Mice were euthanized 1h post-injection, and brain tissue was harvested for immunohistochemical evaluation of microglial activation (Iba1). The proportional area of Iba-1 expression (microglia area %) in the hippocampus and cortex was measured in ImageJ.

Results and Discussion

Dynamic images recorded prior to and after Fe-PyC3A or Gd-DOTA injection into saline and LPS-treated mice demonstrate differential brain signal enhancement in the LPS-treated mice as shown in Figure 1A, B. We found that both SI% and ΔCNR in the hippocampus and cortex at 2 min post-injection of Fe-PyC3A are significantly higher in LPS-treated mice than in saline-treated controls (Figure 1G - J). No significant difference in the SI% and ΔCNR was observed in the hippocampus and cortex between LPS-treated mice and controls using Gd-DOTA (Figure 1C - F). IBA-1 immunostaining demonstrates significantly greater microglial activation in LPS-treated mice vs. controls with 17.3% ± 1.64% vs. 13.4% ± 1.31% (P < 0.001), for the hippocampus, and 25.5% ± 4.34% vs. 21.2% ± 1.40% (P = 0.040) for the cortex (Figure 2A - C). The proportional area of Iba-1 expression significantly and positively correlated with hippocampal SI% (r = 0.880, P < 0.001) and cortex (r = 0.788, P < 0.01) (Figure 3E, F), and ΔCNR (r = 0.640, P = 0.014, Figure 3G), for mice imaged with Fe-PyC3A. In contrast, for mice imaged with Gd-DOTA, no significant correlations between SI% or ΔCNR and Iba-1 positive area were observed in the hippocampus and cortex (Figure 3A-D, H). To further evaluate the specificity of Fe-PyC3A to ROS secreted during neuroinflammation, we compared Iba-1 positive area, Fe-PyC3A induced SI% and ΔCNR in gp91phox knockout (NOX-2 KO) mice. NOX-2 KO mice are deficient in NADPH oxidase II and therefore exhibit severely diminished capabcity to generate ROS. We found that both NOX-2 KO mice and C57BL/6 mice with LPS had greater Iba-1 positive area than C57BL/6 mice with saline (Figure 4). We also found no significant differences between NOX-2 KO mice with LPS and C57BL/6 mice with saline in SI% (P = 0.748) and ΔCNR (P = 0.090), demonstrating the specificity of Fe-PyC3A contrast enhancement for ROS generated through microglial NOX-2 activity.

Conclusion

Fe-PyC3A provided strong, selective contrast enhancement in the hippocampus and cortex of LPS-treated mice. MR imaging using oxidatively activated MR imaging probes such as Fe-PyC3A merits further evaluation as a marker to diagnose and quantify neuroinflammation in neurodegenerative disorders.

Acknowledgements

EMG and IYZ co-corresponding authors. This work was supported by the National Institute on Aging via an Administrative Supplement to R01DK120663. EMG has equity and receivesconsulting income from reveal pharmaceutical.

References

[1] Mi Y, Qi G, Vitali F, et al. Loss of fatty acid degradation by astrocytic mitochondria triggers neuroinflammation and neurodegeneration. Nat Metab. 2023;5(3):445-465.[2] Wang H, Jordan VC, Ramsay IA, et al. Molecular Magnetic Resonance Imaging Using a Redox-Active Iron Complex. J Am Chem Soc. 2019;141(14):5916-5925.