Introduction

Tissue-resident microglia and infiltrated macrophages (M/M) are important mediators of inflammation after ischemic stroke and represent promising therapeutic targets. In the present study, our aim was to investigate the potential of MRI coupled with the injection of a novel nanoparticle (NP), named NanoGd, to image inflammation at the acute stage of ischemic stroke. Our hypothesis was that NanoGd would be phagocytosed by activated M/M. NanoGd-enhanced MRI might thus allow to visualize the brain inflamed areas like iron oxides.^1-3 To understand the biological substrates of NanoGd-induced MR signals, we performed intravital two-photon microscopy back-to-back with MRI in CX3CR1-GFP mice (with fluorescent M/M) submitted to permanent middle cerebral artery occlusion (pMCAO).

Methods

NanoGd NPs are composed of a gadolinium fluoride (GdF3) core coated with bisphosphonate polyethylene glycol (BP-PEG),⁴ functionalized with a near-infrared fluorophore and fully characterized for the purpose of the present study.⁵ In summary, NanoGd has a hydrodynamic diameter of 28 ± 8 nm, a zeta potential of – 42 ± 6 mV, r1 / r2 relaxivities of 0.98 / 20 mM-1.s-1 at 7T (saline, room temperature) and a prolonged vascular remanence (blood half-life ~6h).
To evaluate internalization in-vitro, microglial primary cultures were incubated with NanoGd at [0-1.5 mM] during 24h then imaged with confocal microscopy.
The in-vivo experiment design (Figure 1) was approved by our ethical committee. A total of 31 mice were included. In brief, pMCAo was induced in 28 CX3CR1-GFP mice (sham surgery in 3 mice) at day 0 (D0). Baseline multiparametric MRI included: diffusion-weighted MRI (DWI) and T2-weighted imaging (T2WI) for the assessment of infarct size, T1-weighted imaging (T1WI) pre- and post-Gd-DOTA for the evaluation of blood brain barrier (BBB) integrity and T2*-weighted imaging (T2*WI) for the detection of pre- and post-contrast signal voids. At D1 post-pMCAO, mice were divided into 3 groups: pMCAO (N=6), pMCAO + NanoGd (N=22), sham + NanoGd (N=3). NanoGd was injected intravenously at 2 mmol Gd/kg after baseline MRI. All mice were re-imaged with multiparametric MRI at D3. A subgroup of mice (N=1 pMCAO, N=10 pMCA0 + NanoGd, N=3 sham + NanoGd) underwent two sessions of intravital two-photon microscopy back-to-back with MRI at D1 and D2. At the end of the last imaging session, mice were sacrificed by intracardiac perfusion and their brains were prepared either for x-ray based phase contrast tomography⁶ or for histological analysis. Infarct size and signal voids were quantified longitudinally on MRI. The number of CX3CR1+ cells that have internalized NanoGd were quantified longitudinally on two-photon microscopy. Statistical analyses were performed using (paired) student’s t tests and Pearson correlation analysis. Data are presented as median [25% percentile; 75% percentile].

Results

Confocal microscopy of microglial cultures demonstrated an internalization of NanoGd by M/M (Figure 2) without acute toxicity.
All animals survived and completed the in-vivo protocol. Post-Gd-DOTA T1WI showed BBB disruption in the ischemic lesion of all pMCAO mice at baseline (Figure 3A-B). Post-NanoGd T2WI and T2*WI showed strong hypointense MR signals in the ischemic lesion of pMCAO mice, especially in the ischemic core, not found in control groups (Figure 3C-D). There was only a partial overlap between regions with BBB disruption and those with NanoGd-induced signal voids. Infarct size was stable between baseline and follow-up: 16 [11; 19] mm³ vs 14 [11; 19] mm³ (p=0.968); of note, NanoGd did not alter infarct size at D3 (14 [11; 20] mm³ in pMCAO + NanoGd vs 15 [12; 16] mm³ in pMCAO; p=0.382). In mice injected with NanoGd, the volumes of signal voids were significantly higher at D3 compared to baseline (6 [5; 8] mm³ vs 3 [2; 4] mm³, p=0.0007). In pMCAO + NanoGd, signal voids comprised 47% [36%; 55%] of the ischemic lesion at D3, which was significantly more than in pMCAO without NanoGd (8% [5%-11%], p=0.0001) (Figure 4A). There was no correlation between signal voids extent and infarct size at D3 (R²=0.17). In the subgroup of mice imaged with intravital two-photon microscopy, NanoGd internalization by CX3CR1+ cells was observed in the ischemic lesion (Figure 3F) but not in the extralesional areas (Figure 3E). There was a non-significant trend towards a higher percentage of CX3CR1+/NanoGd+ cells in the ischemic core than in the perilesional area at D1 (73% [69% ; 78%] vs 56% [49% ; 66%], p=0.30) and D2 (72% [64% ; 77%] vs 66% [56% ; 67%], p=0.49). The percentage of CX3CR1+/NanoGd+ cells in the ischemic lesion did not significantly vary between D1 and D2 (63% [59% ; 68%] vs 65% [62% ; 66%], p=0.20, Figure 4B). Post-mortem analyses of perfused brain confirmed the presence of gadolinium in the ischemic lesion (Figure 5A) and the internalization of NanoGd by CX3CR1+ cells (Figure 5B).

Discussion

We here report the first quantitative study using MRI and intravital two-photon microscopy back-to-back to image inflammation in pMCAO thanks to NanoGd, a new multimodal NP. Overall, our data suggest that NanoGd is safe and does not induce inflammation. NanoGd is phagocytosed by M/M both in-vitro and in-vivo. One advantage of NanoGd over iron oxides is the possibility to be imaged specifically with spectral CT.⁴

Conclusion

NanoGd-enhanced MRI may provide a surrogate marker of post-stroke inflammation and might thus be useful for anti-inflammatory treatment evaluation.

Acknowledgements

The authors thank Radu Bolbos and Jean-Baptiste Langlois of the Animage platform (CERMEP, Lyon) for their technical assistance. This research was funded by the French national research agency (ANR) project NanoBrain (grant # ANR-15-CE18-0026-01) and was performed within the framework of the RHU MARVELOUS (ANR16-RHUS-0009) of University Claude Bernard Lyon (UCBL), within the program “Investissements d’Avenir”. We also thank ESRF for allocation of beamtime.