Altered hippocampal microstructure in the epileptogenic rat brain revealed with diffusion MRI using oscillating field gradients
Manisha Aggarwal1, Olli Gröhn2, and Alejandra Sierra2

1Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 2Department of Neurobiology, A. I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland

Synopsis

We investigate changes in the temporal diffusion spectrum sampled using oscillating gradient spin-echo (OGSE) acquisitions at increasing gradient frequencies in the epileptogenic rat brain. PGSE and OGSE data at discrete oscillation frequencies were acquired from pilocarpine-treated and control rat brains (n=5 each) with a spectral resolution of 60 Hz (f = 60 Hz, 120 Hz, 180 Hz). Our findings reveal significant changes in the frequency-dependent modulation of apparent diffusion coefficient (ADC) in specific areas of the pilocarpine brain, which were found to correspond to region-specific gliosis and neuronal loss respectively. Using comparison with histological findings, our results show unique sensitivity of OGSE diffusion MRI to probe specific cellular-level alterations in the epileptogenic brain.

Purpose

The temporal diffusion spectrum in brain tissue reflects sensitivity to restriction effects from intra- to extra-cellular scale structures on free water diffusion1,2. The use of oscillating gradient spin-echo (OGSE) acquisitions allows effective sampling of the spectrum over discrete frequencies, which can provide insights into specific aspects of tissue microstructure, e.g. cell size or density3-6, compared to conventional pulsed-gradient (PGSE) acquisitions which sample the spectrum in the zero-frequency regime. Probing changes in the frequency-dependent characteristics of the temporal diffusion spectrum under pathological conditions could thus potentially provide intrinsic sensitivity to detect cellular-scale alterations in brain tissue. Here, we investigate OGSE-based diffusion contrasts at increasing gradient oscillation frequencies in the rat brain during epileptogenesis. Using comparison with histological findings, we show for the first time, sensitivity of OGSE diffusion MRI to localized gliosis and cellular degenerative changes in the epileptogenic rat brain.

Methods

Adult Wistar rats (n=5) were subjected to status epilepticus (SE) via i.p. injection of pilocarpine. Three weeks after induction of SE, brains of pilocarpine rats (n=5) and age-matched controls (n=5) were perfused with 4% PFA. Imaging experiments were performed on an 11.7T NMR scanner (100 G/cm single-axis gradients) using a 20-mm birdcage transceiver coil. PGSE data of rat forebrains were acquired using 3D gradient-and-spin-echo acquisition (turbo factor/EPI factor = 4/3, TR/TE = 800/44 ms, δ/Δ = 3.2/12 ms, 2 averages) at an isotropic resolution of 120 µm. For OGSE experiments, apodized-trapezoidal oscillating gradient waveforms time-matched to the apodized-cosine waveforms in Parsons et al.7 were implemented, to maximize the b-value and SNR under maximal gradient limits8,9. OGSE data were acquired at three discrete oscillation frequencies with a spectral resolution of 60 Hz (f = 60 Hz, 120 Hz, 180 Hz, and N=1, 2, 3, respectively, with δ=16.67 ms). Tetrahedral diffusion encoding was used with b-value=850 s/mm2 and maximum single-axis gradient of 47 G/cm at 180 Hz. TE for OGSE and PGSE experiments was maintained constant at 44 ms. Maps of apparent diffusion coefficient (ADC) were calculated at each frequency, and the rate of change of ADC with gradient frequency was calculated by linear fitting of the ADC maps versus frequency. Data from each rat were co-registered to one control brain as the reference using diffeomorphic registration. After MRI, the brains were processed for histology, and serially stained using Nissl (thionin) to compare the cytoarchitecture of selected areas.

Results

Fig. 1A-B compares PGSE (0 Hz) and OGSE (180 Hz) ADC maps of control and pilocarpine-treated rat brains. ADC maps at 180 Hz revealed significant (p<0.001) frequency-dependent enhancement of densely-packed cell-layers in the hippocampus of control rats (Fig. 1A’, arrows). No significant differences between control and pilocarpine brains were observed in ADC maps from PGSE experiments (Fig. 1A,B). However, the rate of change of ADC with increasing OGSE frequency was significantly (p<0.005) reduced in the pyramidal cell layer (Py) of pilocarpine rats compared to controls (Fig. 1B’, arrows). Nissl-staining in the same rats (Fig. 1C, D) revealed severe inflammation and neurodegeneration in the densely-packed Py layer of the pilocarpine brain (Fig. 1D).

Interestingly, the cell-sparse lacunosum-moleculare (l-m) layer appears hypointense in fitted ADC versus frequency maps compared to most brain regions, including the cortical grey matter, in control brains (Fig. 1E). However, this subfield was significantly (p<0.005) enhanced in the pilocarpine brains compared to controls at gradient frequencies of 120 Hz and 180 Hz (Fig. 1F, H). Plots in Fig. 1G-H exhibit layer-specific decreases and increases in the frequency-dependent change in ADC in SE brains compared to controls in the Py and l-m, respectively.

PGSE and OGSE maps of the dorsal hippocampus (Fig. 2) revealed drastically different frequency-dependent ADC contrasts in control and pilocarpine brains. OGSE-based ADC maps at 180 Hz exhibited dramatic enhancement of the l-m layer in the pilocarpine brain (Fig. 2B’, arrows), which was not distinguishable in either PGSE contrasts (Fig. 2B) or in control brains (Fig. 2A-A’). Histological assessment (Fig. 2C-D) revealed increased cellular density attributed to gliosis specific to the l-m layer of the pilocarpine brain (Fig. 2D, arrows), which corresponded to the distinctly highlighted layer in the ADC map at 180 Hz (Fig. 2B').

Discussion and Conclusion

Our findings show unique sensitivity of OGSE dMRI to detect cellular-scale microstructural alterations in the epileptogenic rat brain. Both increases and decreases in the frequency-dependence of ADC following induction of SE were observed, and found by histological comparison to correspond to regional gliosis and neuronal loss respectively in specific areas. The range of gradient frequencies here corresponds to root-mean-square-displacements in the range of ~2.8 to 8.2 µm for free water molecular diffusion7,8. Changes in cell morphology and density associated with neurodegeneration and gliosis are complex and involve both proliferation and hypertrophy of glial cells, namely astrocytes and macroglia10. Our findings demonstrate unique sensitivity of frequency-dependent OGSE contrasts to detect these cellular-level processes at specific spatial scales, which could provide key insights into the pathogenesis of epilepsy.

Acknowledgements

National Institutes of Health (NIH), and Academy of Finland

References

1. Stepišnik, J., Time-dependent self-diffusion by NMR spin-echo. Physica B, 1993. 183(4): p. 343-350.

2. Gore, J.C., J. Xu, D.C. Colvin, T.E. Yankeelov, E.C. Parsons, and M.D. Does, Characterization of tissue structure at varying length scales using temporal diffusion spectroscopy. NMR Biomed, 2010. 23(7): p. 745-56.

3. Drobnjak, I., B. Siow, and D.C. Alexander, Optimizing gradient waveforms for microstructure sensitivity in diffusion-weighted MR. J Magn Reson, 2010. 206(1): p. 41-51.

4. Xu, J., M.D. Does, and J.C. Gore, Dependence of temporal diffusion spectra on microstructural properties of biological tissues. Magn Reson Imaging, 2011. 29(3): p. 380-90.

5. Aggarwal, M., M.V. Jones, P.A. Calabresi, S. Mori, and J. Zhang, Probing mouse brain microstructure using oscillating gradient diffusion MRI. Magn Reson Med, 2012. 67(1): p. 98-109.

6. Does, M.D., E.C. Parsons, and J.C. Gore, Oscillating gradient measurements of water diffusion in normal and globally ischemic rat brain. Magn Reson Med, 2003. 49(2): p. 206-15.

7. Parsons, E.C., Jr., M.D. Does, and J.C. Gore, Temporal diffusion spectroscopy: theory and implementation in restricted systems using oscillating gradients. Magn Reson Med, 2006. 55(1): p. 75-84.

8. Ianus, A., B. Siow, I. Drobnjak, H. Zhang, and D.C. Alexander, Gaussian phase distribution approximations for oscillating gradient spin echo diffusion MRI. J Magn Reson, 2013. 227: p. 25-34.

9. Van, A.T., S.J. Holdsworth, and R. Bammer, In vivo investigation of restricted diffusion in the human brain with optimized oscillating diffusion gradient encoding. Magn Reson Med, 2014. 71(1): p. 83-94.

10. Represa, A., J. Niquet, H. Pollard, and Y. Ben-Ari, Cell death, gliosis, and synaptic remodeling in the hippocampus of epileptic rats. J Neurobiol, 1995. 26(3): p. 413-25.

Figures

Fig. 1: A-B) Comparison of ADC maps of representative control (A-A’) and pilocarpine-treated (B-B’) rat brains from PGSE (0 Hz) and OGSE (180 Hz) acquisitions. C-D) Nissl-stained sections from the same rats showing the pyramidal (Py) layer. E-F) Fitted ADC versus frequency maps averaged over control and SE cohorts (n=5 each). G-H) Plots of ADC versus gradient-frequency for Py and l-m layers (mean ± standard deviation, for n=5 rats per group).

Fig. 2: A-B) PGSE and OGSE (f=180 Hz) ADC maps of the dorsal hippocampus revealed significantly different frequency-dependent contrasts in the control (A-A’) and pilocarpine-treated (B-B’) rats. C-D) Nissl-stained sections from the same rats indicate regional gliosis in the l-m layer of the pilocarpine brain (arrows in D), corresponding to the significantly enhanced layer in the ADC map at 180 Hz (arrows in B’), which is not seen in the control brain (A’).



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
1083