1211

Iron-Induced MR Contrast in Human Locus Coeruleus: A Cautionary Tale of Misleading Post Mortem MRI Results
Evgeniya Kirilina1,2, Charlotte Lange1,3, Carsten Jäger1, Tilo Reinert1,3, Kerrin Pine1, Thomas Lohmiller4, Siawoosh Mohammadi1,5, Tobias Streubel1,5, Malte David Brammerloh1,3, Anneke Alkemade6, Birte Forstmann6, Andreas Herrler7, Alexander Schnegg8, Markus Morawski9, and Nikolaus Weiskopf1,3

1Department of Neurophysics, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 2Neurocomputation and Neuroimaging Unit, Department of Education and Psychology, Free University Berlin, Berlin, Germany, 3Felix Bloch Institute for Solid State Physics, Faculty of Physics and Earth Sciences, Leipzig University, Leipzig, Germany, 4Berlin Joint EPR Lab, Institute for Nanospectroscopy, Helmholtz-Zentrum Berlin fuer Materialien und Energie, Berlin, Germany, 5Department of Systems Neurosciences, University Medical Center Hamburg-Eppendorf, Hamburg, Germany, 6Integrative Model-Based Neuroscience Research Unit, University of Amsterdam, Amsterdam, Netherlands, 7Department of Anatomy and Embryology, Faculty of Health, Medicine and Life Science, Maastricht University, Maastricht, Netherlands, 8EPR Research Group, Max Planck Institute for Chemical Energy Conversion, Mülheim, Germany, 9Paul Flechsig Institute of Brain Research, Leipzig University, Leipzig, Germany

Synopsis

MR contrast mechanisms in human locus coeruleus were studied combining high-resolution post mortem MRI, histology, ion-beam microscopy, and electron paramagnetic resonance. We demonstrate that the main source of MR contrast in formalin fixed LC is paramagnetic iron accumulated in noradrinergic neurons. However, we show that MR contrast in LC drastically changes during the first six months of tissue fixation. We assign these changes to iron been scavenged by neuromelanin and the change of its paramagnetic state. The results have major consequences for MRI of the locus coeruleus, demonstrating a fundamental change rather than the commonly known gradual changes in contrast due to formalin fixation.

INTRODUCTION

The locus coeruleus (LC), a small nucleus in the pons1, is affected in early stages of Alzheimer's disease (AD)2–7. MRI promises much needed in vivo biomarkers of LC integrity, for early AD diagnosis and treatment monitoring. LC visualisation and quantification8–10 utilize MR contrasts, induced by the pigment and metal chelate neuromelanin (NM) contained in noradrenergic (NA) neurons in LC10. Yet, for quantitative markers and a mechanistic link between MRI and the neuropathology in LC, a fundamental understanding of MRI contrast mechanisms is indispensable. MR contrast mechanisms in LC are still not understood and differ from those in grey and white matter as well as in other NM containing nuclei like the substantia nigra (SN) pars compacta. Particularly the influence of iron absorbed in NM on proton relaxation rates is unknown. For example R2* hyperintensity observed in SN was not detected in LC in vivo. We combine high-resolution post mortem MRI, histology/immunocytochemistry, ion-beam microscopy and electron paramagnetic resonance (EPR) for comprehensively studying the contrast mechanisms in LC. We demonstrate that the main source of MR contrast in formalin fixed LC is paramagnetic iron accumulated in NM-containing neurons. Furthermore, we show that MR contrast in LC drastically changes during the first six months of tissue fixation. We assign these changes to iron been scavenged by NM and the change of its paramagnetic state.

METHODS

Six human post mortem whole-brain specimens and six tissue blocks, fixed in paraformaldehyde for periods between 9 days and 3 years, all encompassing bilateral LC, were investigated by 7T MRI (Magnetom, Siemens Germany). Effective transverse relaxation rate (R2*) maps were recorded using a multi-echo FLASH acquisition11 (repetition time TR=180 ms, echo times TE1-16=2.4-40 ms, resolutions 0.4 and 0.2 mm for whole brains and tissue blocks, respectively). Transverse relaxation rate (R2) maps were acquired using a spin-echo sequence. High resolution T2*w images (resolution 50 μm) were acquired for the tissue blocks. The contribution of iron to the R2 and R2* was quantified, by using chemical iron extraction on one tissue block12. Quantitative iron concentration maps were obtained using Proton-Induced X-ray Emission (PIXE, proton beam LIPSION13). LC from two tissue blocks with short (14 days) and long (160 days) fixation times were dissected and investigated using X-band EPR for characterisation and quantification of the paramagnetic metals in NM14.

RESULTS AND DISCUSSION

Increased values of R2* were observed in the LC only in tissue samples with prolonged fixation times ( >5 months) (Figure 1). T2*w images showed dot-like low intensity structures at the positions corresponding to the locations of iron-rich NA neurons in LC (Figures 2, 3c) indicating NA neurons as the cause of enhanced R2*. We found that hyperintensity in R2* and R2 as well as the dot-like structures in T2*w images disappeared after iron extraction (Figure 3), in line with the primary contrast source being iron contained in NA neurons. Iron-induced relaxation rates ΔR2*=52±10 s-1 were found to be much higher than ΔR2=12±10 s-1, in line with the dominating contribution of static de-phasing15 mechanisms to R2* (Figure 3c). Combined with a quantitative iron estimation from PIXE (Figure 3d) the relaxivity of NM-bound iron was determined (r2*= 4.9 s-1/ppm, r2=1.12 s-1/ppm). The R2* contrast between LC and the surrounding tissue significantly increased within the first months of fixation (Figure 4). While no significant difference in R2* between LC and the surrounding tissue was found for fixation time of 9 days (5±10 s-1), a pronounced contrast (25±10 s-1) was observed in all samples that were fixed for long times. EPR spectra of LC samples revealed the presence of NM-bound16 paramagnetic Fe3+ (ground spin state S=5/2, g=4.3, rhombic ligand coordination) in concentrations 0.87±0.1 (14d fixation) and 3.2±0.3 10-5 M (160d fixation). The 3.7-fold increase in paramagnetic iron concentration due to fixation indicates either saturation of LC NM with iron released from tissue due to fixation or a change of the paramagnetic state of NM iron.

CONCLUSIONS

Paramagnetic iron in NM in NA neurons is the primary source of R2* and R2 contrast in fixed post mortem LC tissue. The intracellular iron concentration in NM neurons was quantified and quantitatively linked to the observed R2* and R2 maps. Iron-induced contrast in LC drastically changes during the first months of fixation. This can be either assigned to iron accumulation in unsaturated NM in LC, or to a modulation of the iron oxidation and spin state. The results have major consequences for MRI of the LC, demonstrating a fundamental change rather than the commonly known gradual changes in contrast due to formalin fixation. Since the in vivo and post mortem MRI cannot readily be compared, histological validation studies and developing AD biomarkers based on the LC are complicate.

Acknowledgements

We thank the Brain Banking Centre Leipzig of the German Brain-Net, operated by the Paul Flechsig Institute of Brain Research, (Medical Faculty, University of Leipzig, Department of Neuropathology, University Hospital Leipzig), the Department of Legal Medicine, Medical Center Hamburg-Eppendorf and the body domantion program operated by the Department of Anatomy and Embryology at Maastricht University for providing post mortem tissue. The entire procedure of case recruitment, acquisition of the patient's personal data, the protocols and the informed consent forms, performing the autopsy, and handling the autopsy material have been approved by the responsible authorities (Approval by the Sächsisches Bestattungsgesetz von 1994, 3. Abschnitt, §18, Ziffer 8; GZ 01GI9999-01GI0299; Approval \# WF-74/16, Approval \# 82-02 and Approval \# 205/17-ek, Approval \# 153/17-ek).

The research leading to these results has received funding from the European Research Council under the European Union's Seventh Framework Program (FP7/2007-2013) / ERC grant agreement n° 616905. This project has also received funding from the BMBF (01EW1711A & B) in the framework of ERA-NET NEURON.

M. D. Brammerloh has received funding from the International Max Planck Research School on Neuroscience of Communication: Function, Structure, and Plasticity.

References

1. Fernandes, P., Regala, J., Correia, F. & Gonçalves-Ferreira, A. J. The human locus coeruleus 3-D stereotactic anatomy. Surg. Radiol. Anat. 34, 879–885 (2012).

2. Benarroch, E. E. Brainstem in multiple system atrophy: clinicopathological correlations. Cell. Mol. Neurobiol. 23, 519–526 (2003).

3. Braak, H. et al. Staging of brain pathology related to sporadic Parkinson’s disease. Neurobiol. Aging 24, 197–211 (2003).

4. Eser, R. A. et al. Selective Vulnerability of Brainstem Nuclei in Distinct Tauopathies: A Postmortem Study. J. Neuropathol. Exp. Neurol. 77, 149–161 (2018).

5. Arendt, T., Brückner, M. K., Morawski, M., Jäger, C. & Gertz, H.-J. Early neurone loss in Alzheimer’s disease: cortical or subcortical? Acta Neuropathol. Commun. 3, 10 (2015).

6. Rüb, U. et al. The Brainstem Tau Cytoskeletal Pathology of Alzheimer’s Disease: A Brief Historical Overview and Description of its Anatomical Distribution Pattern, Evolutional Features, Pathogenetic and Clinical Relevance. Curr. Alzheimer Res. 13, 1178–1197 (2016).

7. Ehrenberg, A. J. et al. Quantifying the accretion of hyperphosphorylated tau in the locus coeruleus and dorsal raphe nucleus: the pathological building blocks of early Alzheimer’s disease. Neuropathol. Appl. Neurobiol. 43, 393–408 (2017).

8. Priovoulos, N. et al. High-resolution in vivo imaging of human locus coeruleus by magnetization transfer MRI at 3T and 7T. NeuroImage 168, 427–436 (2018).

9. Keren, N. I., Lozar, C. T., Harris, K. C., Morgan, P. S. & Eckert, M. A. In vivo mapping of the human locus coeruleus. NeuroImage 47, 1261–1267 (2009).

10. Keren, N. I. et al. Histologic validation of locus coeruleus MRI contrast in post-mortem tissue. NeuroImage 113, 235–245 (2015).

11. Weiskopf, N. et al. Quantitative multi-parameter mapping of R1, PD(*), MT, and R2(*) at 3T: a multi-center validation. Front. Neurosci. 7, 95 (2013).

12. Stüber, C. et al. Myelin and iron concentration in the human brain: A quantitative study of MRI contrast. Neuroimage 93, 95–106 (2014).

13. Reinert, T., Fiedler, A., Morawski, M. & Arendt, T. High resolution quantitative element mapping of neuromelanin-containing neurons. Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. At. 260, 227–230 (2007).

14. Shima, T. et al. Binding of iron to neuromelanin of human substantia nigra and synthetic melanin: an electron paramagnetic resonance spectroscopy study. Free Radic. Biol. Med. 23, 110–119 (1997).

15. Yablonskiy, D. A. & Haacke, E. M. Theory of NMR signal behavior in magnetically inhomogeneous tissues: the static dephasing regime. Magn. Reson. Med. 32, 749–763 (1994).

16. Zecca, L. et al. The role of iron and copper molecules in the neuronal vulnerability of locus coeruleus and substantia nigra during aging. Proc. Natl. Acad. Sci. U. S. A. 101, 9843–9848 (2004).

Figures

Figure 1. Quantitative R2* maps of post mortem brain samples after short (top row), intermediate (middle row) and long (bottom row) fixation times. The position of the bilateral LC in the axial slice is indicated by arrows. Images for whole brain samples (b, c, g, h, i) were acquired with an isotropic resolution of 0.4 mm using a 32-channel RF head coil. Images for tissue blocks (a, d, e, f) were acquired with an isotropic resolution of 0.2 mm using a custom-made 2-channel transmit-receive RF coil. The LC is clearly visible in all samples showing higher R2* after 5 months of fixation.

Figure 2. NM-containing NA neurons are the main source of R2* hyperintensity in LC. (a) Axial slice of an ultra-high resolution T2* weighted image acquired on a tissue block containing LC (50 μm isotropic resolution, TE=19 ms, 169 days of fixation). NA neurons are visible in LC as shown by an arrow. In addition, myelinated fiber tracts around LC are visible; (b) Tyrosine hydroxylase immunoreactivity for NA neurons of the same slice. NA neurons are visible; (c) Luxol Fast Blue stain for myelinated fibers shows the location of dense fibre tracts surrounding LC. NM-containing neurons are visible;

Figure 3. Iron in NM dominates R2*. Quantitative R2* (a) and R2 (b) maps of tissue block before (left part) and after (right part) chemical iron extraction from tissue. (c) R2* and R2 values averaged over LC before and after iron extraction. Higher iron-induced contributions to R2* indicate the predominant contribution of a static de-phasing mechanism. (d) Quantitative maps of iron concentration measured with PIXE in LC before and after chemical iron extraction. High iron concentrations were observed in NA neurons. Averaged iron concentration in LC was found to be 10.7±2 mm/g wet tissue weight.

Figure 4. R2* contrast between LC and the surrounding tissue as a function of fixation time in formalin. Difference of averaged R2* between LC and surrounding tissue are plotted for 12 investigated samples. Bars represent error of mean. Dashed line shows sigmoid function fitted to the data.

Figure 5. X-band EPR spectra of LC samples with different fixation times, recorded at 70K. Spectral components corresponding to iron and copper are marked with arrows and superimposed by simulations. Total spin concentrations for both species were determined via the double integrals of their EPR spectral components, calibrated against a 1mM CuSO4 standard with known spin count and weighted with respect to their transition probabilities.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
1211