Back to Projects

Emerging Imaging

Translating brain magnetic resonance imaging signals to iron and myelin to appraise Alzheimer’s disease

Project ID: 2021_030

1st Supervisor: Po-Wah So, King’s College London
2nd Supervisor: Mads S Bergholt, King’s College London
Additional Supervisor: Shaihan Malik, King’s College London
Clinical Champion: Dag Aarsland, King’s College London

Aim of the PhD Project:

To determine the individual and combined contributions of iron and myelin to multi-modality quantitative magnetic resonance imaging (QMRI) signals.

Project Description / Background:

Iron accumulates in the brain with ageing [1], with advancing age being the major risk factor for neurodegenerative diseases such as Alzheimer’s Disease (AD; [2]). Indeed, iron dyshomeostasis is a feature of AD [3] and iron chelation therapy is undergoing a clinical trial for AD.

Quantitative magnetic resonance imaging MRI (QMRI) methods are sensitive to iron content (to differing extents and specificities). The So Lab and others have shown T2* values compared to T1 and T2, correlated best with iron [1,4]. Unusually, the So Lab has correlated QMRI data with quantitative spatial iron measurements obtained by synchrotron radiation X-ray fluorescence (SRXRF), rather than rely on quantitative bulk iron analyses or non-quantitative histochemical iron staining.

Myelin is also known to significantly modulate QMRI signals [5,6] and the situation is further complicated by the high iron content of myelin itself. Myelin is formed from the wrapping of oligodendrocyte membranes around axons and functions as “electrical insulation” to ensure fast nerve conduction. Conventionally, myelin is assessed by (immuno)histochemical staining and qualitative, rather than quantitative.

Generally, relationships between QMRI with iron and/or myelin are evaluated by ex vivo QMRI of brain samples and then sectioning of the sample for correlative iron and/or myelin histology as mentioned above. Disparate resolutions between such diverse data types and artefacts from histological processing contributes to inaccuracies/ambiguities, especially when co-registering datasets. Using bespoke high signal-to-noise MRI coils (previously developed by the So lab with PulseTeq Ltd), thin sections of brain tissues will undergo high resolution QMRI prior to quantitative iron and myelin mapping by SRXRF/laser-ablation-inductive coupled plasma-mass spectrometry (LA-ICP-MS) [1,7] and desorption electrospray ionisation-mass spectroscopic (DESI-MSI)/Raman imaging [8], respectively. Uniquely, multi-modality imaging will be performed at comparable resolutions with minimal sample displacement between modalities. High resolution QMRI of thin brain samples is challenging but made possible using bespoke MRI coils. In this manner, registration errors between datasets are minimised and determination of accurate pixel-wise relationships between individual QMRI, iron and myelin imaging datasets are possible. Notably, lipid composition and myelin structural information can also be obtained to determine relationships between QMRI and specific lipid types/myelin structure.

In this project, we aim to correlate iron- and myelin-sensitive QMRI signals with quantitative state-of-the-art physiochemical iron and myelin mapping developed by So and Bergholt Labs [1,8], respectively. Age-matched control and AD brain samples will be obtained from the Brain Bank for Dementia (which Dr So has previously accessed) and analysed, for potential future translation to monitoring iron chelation therapies in man. While AD is often considered a grey matter disease, white matter myelin has also been shown to be deranged [4]. Teasing apart contributions of myelin and iron to multimodality QMRI measurements aids true assessment of the roles of iron dyshomeostasis and myelin dysfunction in AD for identification of novel AD therapeutics and monitoring.

Applicants will ideally have knowledge and experience in (analytical) chemistry, physics, or (biomedical) engineering, and possibly in biology/neuroscience.

Figure shows a typical workflow for the multi-modality correlative imaging of brain samples (of age-matched control and Alzheimer’s disease), encompassing quantitative MRI and physico-chemical quantitative mapping of iron and myelin.

Figure 1: Figure shows a typical workflow for the multi-modality correlative imaging of brain samples (of age-matched control and Alzheimer’s disease), encompassing quantitative MRI and physico-chemical quantitative mapping of iron and myelin.

References

  1. Walker T, Michaelides C, Ekonomou A, Geraki K, Parkes HG, Suessmilch M. Herlihy A, So P-W (2016). Aging (Albany NY) 8(10): 2488.
  2. Ashraf A, Clark M, So P-W (2018). Fronts Aging Neursci. 10: 65.
  3. Ashraf A, Jeandriens J, Parkes HG, So P-W (2020). Redox Biol 32: 101494.
  4. Bulk M, Abdelmoula WM, Nabuurs RJA, van der Graaf LM, Mulders CWH, Mulder AA, et al. (2018) Neurobiol Aging, 62: 231.
  5. Heath F, Hurley SA, Johansen-Berg H, Sampaio-Baptista C (2018). Developmental Neurobiol DOI 10.1002/dneu.22552.
  6. Varma G, Duhamel G, de Bazelaire C, Alsop DC (2015). MRM 73(2): 614.Harkins KD, Xu J, Dula AN, Li K, Valentine WM, Gochberg DF, et al. 2016. MRM 75(3): 1341.
  7. O’Reilly J, Douglas D, Braybrook J,  So P-W, Vergucht E, Garrevoet J, Vekemans B, Vincze L, Goenaga-Infante H (2014). JAAS 28(8): 1374.
  8. Bergholt MS, Serio A, McKenzie JS, Boyd A, Soares RG, Tillner J, Chiappini C, Wu V, Dannhorn A, Thakats Z, Williams A, Stevens MM (2018). ACS Central Science 4: 39.

Back to Projects