4322
Non-invasive detection of toxin-induced inflammation in the mouse brain using hyperpolarized 13C MRSI1Physical Therapy and Rehabilitation Science, University of California, San Francisco, San Francisco, CA, United States, 2Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, United States
Synopsis
In recent years, hyperpolarized (HP) 13C magnetic resonance spectroscopic imaging (MRSI) has shown promise in assessing neuroinflammation in the brains of mouse models of multiple sclerosis and traumatic brain injury. In this follow-up study, we used 13C MRSI of HP [1-13C] pyruvate/13C urea to assess neuroinflammation following intracranial injection of a toxin, namely lipopolysaccharide (LPS). A significantly elevated lactate:pyruvate ratio was observed in LPS-injected mice, and was associated with a significant increase in the presence of macrophages/microglia and astrocytes as assessed histologically. Our results demonstrate that HP 13C MRS has the sensitivity to assess toxin-induced changes in the brain.
Introduction
Neuroinflammation is a key player in many neurodegenerative diseases1. Non-invasive methods to monitor neuroinflammation are critically needed. To date, a few studies have demonstrated the potential of 13C magnetic resonance spectroscopic imaging (MRSI) of hyperpolarized (HP [1-13C] pyruvate to monitor chronic neuroinflammation in murine models of multiple sclerosis2 and traumatic brain injury3. Here, we aimed to establish whether HP [1-13C] pyruvate would be sensitive to metabolic changes caused by toxin-induced neuroinflammation, specifically after lipopolysaccharide (LPS) injection4,5. Using immunofluorescence, we measured increased numbers of microglia at 7 days after LPS injection, and increased astrocytes at 3 and 7 days. We subsequently observed that values of HP lactate:pyruvate normalized to urea were significantly increased at 7 days. This demonstrates the potential of 13C MRSI of HP [1-13C] pyruvate to monitor LPS-induced neuroinflammation.Methods
Animals: 20 C57BL6/J male and female mice (Figure 1A, histology: n=15; MR imaging: n=13) underwent intracranial injections of LPS or saline into the right striatum (ipsilateral side) using a stereotaxic frame (coordinates 1.55:1.55:2.8, Stoelting, IL), under 1-2% isoflurane anesthesia.
Histological characterization: For each time point, animals were transcardially perfused and brains fixed (4% PFA); n=5 baseline (prior to injection), n=3 24 hours post-surgery (day 1), n=6 three days post-surgery (day 3), and n=6 seven days post-surgery (day 7). Fluorescence images of 10μm tissue slices were acquired following Iba1 and GFAP staining.
MR acquisitions: Thirteen C57BL/6J mice (n=7 LPS, n=6 saline) were imaged at baseline, day 3 and day 7 on a 3 Tesla horizontal system (Bruker) with a dual-tuned 1H/13C mouse head coil (2cm diameter). T2-weighted MRI was acquired (FOV 20x20mm, 192x192 matrix, NA=4, TR=2s, TE=60ms, 9 slices, thickness 1mm). 24μl [1-13C] pyruvate and 55μl 13C urea were co-polarized for ~1h in a Hypersense polarizer (Oxford Instruments), and rapidly dissolved in 4.5ml heated buffer (80mM NaOH in PBS) of which 300μl was injected via a tail vein catheter over 14s; data were acquired every 4.2s using a dynamic 2D-CSI sequence (slice thickness 5mm, FOV 24x24, 8x8 matrix, TR=66.4ms, TE=1.24ms) (Figure 1B).
13C MRS data analysis: Ipsilateral and contralateral voxel data (Figure 1C) were analyzed after voxel-shift post-processing using in-house SIVIC software and custom-built MATLAB scripts. Dynamic spectra from each voxel were summed (Figure 1D) and peaks fitted to a Lorentzian curve prior to integration to obtain lactate, pyruvate and urea values.
Statistical analysis: Results are expressed as mean ± standard deviation. Statistical tests were carried out as follows (with appropriate multiple comparisons corrections): microscopy data: two-way ANOVA; longitudinal in vivo data between timepoints: repeated measures one-way ANOVA; in vivo data normalized to contralateral: two-way ANOVA.
Results
Immunofluorescence data confirmed the expected effect of the LPS toxin4,5, showing a significantly elevated number of Iba1-positive cells in the LPS-injected side compared to contralateral at day 7 (p<0.0001, Figure 2A&C). GFAP-positive cells were similarly significantly increased at day 3 and day 7 in the LPS-injected side compared to contralateral (p=0.02, 0<0.0001 respectively, Figure 2B&D).
HP pyruvate, lactate and urea peaks were clearly visible in all voxels in each group (Figure 3A). Normalization to the contralateral side of the brain was first performed to enable comparisons between groups at all timepoints (Figure 3B). Normalized lactate:pyruvate ratios in the LPS animals were significantly increased at day 7 compared to day 3 and baseline, and also in comparison to the day 7 saline data, in line with increased Iba1 and GFAP staining at that time point.
LPS-treated animals showed increased ipsilateral lactate:pyruvate ratios at day 7 compared to day 3, which was in turn higher than baseline (Figure 4A). This group also showed elevated contralateral ratios at day 3 and day 7 compared to baseline, potentially due to a diffuse effect of LPS throughout the brain, but no differences were seen between day 3 and day 7. Importantly, no differences were observed in the saline-injected group, confirming effects observed were not a result of surgery. On normalization to 13C urea, day 7 lactate:pyruvate ratios in the LPS-injected side were significantly higher compared to day 3 or baseline (Figure 4B). Elevated contralateral ratios were also seen at day 7 compared to day 3. The saline-injected group again showed no differences.
Conclusion
We have successfully observed the response of the brain to LPS using HP [1-13C] pyruvate and 13C urea. We showed that the increased lactate:pyruvate ratios at seven days is in line with elevated numbers of Iba1 and GFAP-positive cells, and with previous reports2,3,6, thus demonstrating the potential of HP 13C MRSI in assessing neuroinflammation in a wide variety of diseases.Acknowledgements
This work has been supported by: NIH R01NS102156, Cal-BRAIN 349087, NMSS research grant RG-1701-26630, NMSS fellowship FG-1507-05297, the Hilton Foundation #17319, the NIH Hyperpolarized MRI Technology Resource Center #P41EB013598 and the Dana Foundation. The authors would like to thank Dr. Chloe Najac for invaluable help with MATLAB scripts.References
1. Hickman, S., Izzy, S., Sen, P., Morsett, L. & El Khoury, J. Microglia in neurodegeneration. Nat. Neurosci. doi:10.1038/s41593-018-0242-x
2. Guglielmetti, C. et al. Hyperpolarized 13 C MR metabolic imaging can detect neuroinflammation in vivo in a multiple sclerosis murine model. Proc. Natl. Acad. Sci. 201613345 (2017). doi:10.1073/pnas.1613345114
3. Guglielmetti, C. et al. In vivo metabolic imaging of Traumatic Brain Injury. Sci. Rep. 7, 17525 (2017).
4. Go, M., Kou, J., Lim, J.-E., Yang, J. & Fukuchi, K. Microglial response to LPS increases in wild-type mice during aging but diminishes in an Alzheimer’s mouse model: Implication of TLR4 signaling in disease progression. Biochemical and Biophysical Research Communications 479, (2016).
5. Herber, D. L. et al. Time-dependent reduction in Aβ levels after intracranial LPS administration in APP transgenic mice. Exp. Neurol. 190, 245–253 (2004).
6. DeVience, S. J. et al. Metabolic imaging of energy metabolism in traumatic brain injury using hyperpolarized [1-13C]pyruvate. Sci. Rep. 7, 1907 (2017).
Figures
