In our lab, we aim to unravel the molecular underpinnings of normal and pathological cognitive aging in the human brain by means of multimodal neuroimaging. Therefore, we apply different neuroimaging modalities, such as functional and structural MRI (Magnetic Resonance Imaging) at (ultra-)high field strengths of 3 and 7 Tesla, and PET (Positron Emission Tomography).
Aging is often associated with the accumulation of tau tangles and Amyloid-beta (Aβ plaques), the hallmarks of AD, but also with vascular pathology. Notably, these pathologies affect certain regions, with tau accumulating in the anterior temporal lobe or Aβ in posterior-midline regions (e.g. Maass, et al., 2017, NeuroImage) . In the progression from normal aging towards Alzheimer’s disease (AD), these proteins spread throughout the brain via distinct networks (e.g. Adams, et al., 2019, eLife). This accumulation of pathology is associated with early regional and network dysfunction, before neurodegeneration and cognitive impairment becomes apparent. Interestingly, this early accumulation of pathology also relates to memory deficits even in otherwise cognitively unimpaired individuals (e.g. Maass, Berron et al, 2019, Brain)and likely interact in the progression from normal aging to AD. By combining molecular and MR imaging of the brain, we can study how age-related pathologies emerge and how this affects brain function. Specifically, PET allows to measure regional patterns of tau (Schöll*, Maass* et al., 2019, Mol. Cell. Neurosci.) and Aβ accumulation, and Glucose metabolism, which is not possible with fluid biomarkers. FMRI allows us measure regional dysfunction, before cognitive impairment and profound neurodegeneration emerges (Billette*, Ziegler* et al., 2022, Neurology). By using this multimodal neuroimaging approach, our group aims to understand why some individuals accumulate age-related pathology, why others age successfully (are resistant), and which beneficial or detrimental factors might underlie this vulnerability). We further aim to understand why some individuals who develop pathology still remain cognitively healthy (i.e. are resilient). This approach will help to develop interventions that promote healthy aging and prevent cognitive decline (e.g. Maass et al, 2015, Mol Psychiatry). In this respect, we want to better understand which factors underlie or impact brain plasticity. Positioned at the interface of basic cognitive and molecular neuroscience, advanced multimodal neuroimaging, and clinical application, our research integrates knowledge from different disciplines to understand the molecular underpinnings of healthy and pathological aging.
Several fMRI studies in humans have found increased brain activity in hippocampal and posterior midline regions, including the precuneus, in the presence of early Aβ and tau burden (e.g. Mormino et al., 2011; Sperling et al., 2009)(e.g. Maass, Berron et al., 2019, Brain ;Adams, et al., 2020, JNeurosci). These findings are in line with animal models in which Aβ or tau pathology has been linked to higher neural excitability. However, it is still unclear whether the increased fMRI activity in humans is compensatory for pathology, driven by pathology or even driving the accumulation of Aβ or tau itself. Moreover, the contribution of other factors such as neuroinflammation or vascular changes to fMRI BOLD changes remains unclear.
Recently, our group examined how fMRI-task activity during novelty processing differs across the AD risk spectrum using the DELCODE study (Jessen et al., 2018). We found an inverted-U shape pattern of activation across the AD risk spectrum (Billette*, Ziegler* et al, 2022). Specifically, we observed increased activity in patients with subjective cognitive decline (SCD) or mild cognitive impairment (MCI), which was reduced in AD dementia relative to MCI. However, precuneus activity was not related to AD biomarkers, APOE4 genotype or brain volume. While this suggests that increased precuneus activity might represent an early signature of memory impairment, the underlying mechanism remains unclear. In future studies we will investigate what underlies and causes the increased fMRI activity that we observed in SCD and MCI. In this respect, we collaborate with Prof. Boecker’s group at DZNE Bonn and Prof. Schott’s group at DZNE Göttingen.
To determine whether this increased fMRI activation is (i) caused by or (ii) directly driving protein accumulation or (iii) whether it reflects compensatory processes for AD pathology in humans, longitudinal data and translational research is required. In a collaborative study with Prof. Sylvia Villeneuve on the Prevent AD Cohort established in Montreal, we are currently assessing how brain activity during memory encoding and retrieval changes over time in cognitively normal older adults at risk for AD. We further study how these activity alterations are related to changes in AD pathology measured by CSF and/or PET (Molloy et al, AAIC 2022).
Neuroinflammation is a hallmark of AD and other neurodegenerative disorders and is likely to be triggered throughout different stages of disease. For example, this can happen via pathogenic protein aggregates or in response to neuronal death. In this respect, microglial activation might be initially beneficial and mediate phagocytic clearance of Aβ, whereas chronic neuroinflammation is thought to trigger the release of proinflammatory cytokines and contribute to disease progression and severity. Yet, the exact time course of this response and the effect on local and global brain structure and brain function in the course of AD remains to be explored, particularly in early disease stages.
In a collaboration with Prof. Michael Heneka’s group and the DZNE Bonn, we showed that certain inflammatory markers in cerebrospinal fluid (CSF) are associated with preserved brain structure and cognitive function in the DZNE DELCODE cohort (Brosseron*, Maass*, Kleineidam* et al., 2022, Neuron). We showed that CSF levels of several inflammation-related markers were elevated in healthy subjects with pathological levels of tau or other neurodegeneration markers. Intriguingly, higher CSF levels of soluble TAM receptors sTyro and sAXL were related to higher brain volume/thickness, better cognition and less decline at follow-up. Our findings indicate a protective mechanism of TAM (Tyro3, AXL, Mertk) receptor signaling, in line with its role in regulation of microglial activation, phagocytosis and pathological protein aggregate clearance (e.g. Tondo et al., 2019, Dis Markers; Huang et al., 2021, Nature Immunology).
Following these findings, our group further assessed whether there are beneficial or detrimental “inflammatory signatures“ (or groups of inflammatory markers) in terms of their relationship to brain integrity and cognitive function in the DELCODE cohort (Hayek et al., AAIC 2022). Our data suggest there are different inflammatory signatures (reflecting different processes), where some markers such as TAM receptors or sTREM2 might reflect a damage response towards aging and AD that is, however, beneficial or protective. In contrast, an inflammatory signature of increased proinflammatory cytokines or CRP might reflect proinflammatory processes that are detrimental to brain health and cognition.
Episodic memory performance declines at approximately t age 60 in most healthy individuals . Medial temporal lobe tau accumulation is commonly present even in the presence of this “normal cognitive aging”, and can thus be considered a trait of normal aging . However, higher medial temporal lobe tau burden relates to reduced memory performance and to gray matter atrophy in clinically unimpaired individuals. When assessing regional AD pathology and atrophy, we previously found that medial temporal lobe tau pathology (measured with PET) explained 20% of the variance in episodic memory performance (e.g. Maass, et al., 2017, JNeurosci) . However, most of the variance remained unexplained. This suggests that some individuals may be able to tolerate or respond to pathology, thereby minimizing its impact on cognition (see Arenaza-Urquijo and Vemuri., 2018, Neurology). Why do some individuals perform better than predicted by their pathological burden(are resilient) and why do others not accumulate age-related pathology (are resistant)?
While some memory decline in old age is “normal” , there are also older individuals aged 80 years whose performance matches that of individuals 20 to 25 years their junior. These are commonly referred to as "SuperAgers" (e.g. Harrison, et al, 2018, Neurobiol Aging). How do the brains of these older adults, who perform well above expected for their age, differ in terms of pathology (e.g. tau burden or vascular pathology) and neural resources (e.g. brain volume, neural activity, connectivity, blood flow)? Which environmental or genetic factors relate to superior cognition? These are major questions that we study within the DFG-funded Collaborative Research Center CRC1436 “Neural Resources of Cognition” (first funding period 2021-2024). The overarching goal of the CRC is to uncover the performance limits of the human brain and to explore methodological approaches for improving this performance with targeted interventions. More than 40 scientists in 22 individual projects are currently working in this CRC.
Within the CRC1436, we are building a molecularly characterized aging cohort, as part of the central project Z03 that is co-lead by Prof. Emrah Düzel and Prof. Michael Kreissl: sfb1436.de/projects/human-molecular-imaging-ageing-and-superageing-cohort/. We are currently recruiting cognitively healthy older adults (≥ 60 years) including SuperAgers, to study the molecular, functional, and structural characteristics of successful cognitive aging. Participants will be characterized as SuperAgers if they are aged 80 years or older and their memory performance is at or above the average normative values for healthy individuals in their 50ies to 60ies. All participants will undergo a 3 T MRI scan to assess brain structure, brain function and blood flow. A subsample of subjects will undergo MR-PET scans to measure regional tau accumulation via the novel second-generation tracer PI-2620 from Life Molecular Imaging. Tau PET will be performed in collaboration with the Nuclear Medicine Department in Leipzig led by Prof. Osama Sabri, while blood plasma will be used for detection of AD biomarkers in collaboration with the DZNE Göttingen (Prof. Jens Wiltfang).
People who are interested to participate are always welcome and can register here:
In collaboration with several other groups in Magdeburg (Prof. Schreiber, Prof. Düzel, Prof. Speck), our recent research suggests that the vascularization pattern of the hippocampus might be one important factor that mediates brain reserve in aging. Until recently, hippocampal vascularization could only be studied post-mortem and these ex vivo studies showed that all hippocampal arteries originate from two major arteries: the posterior cerebral artery (PCA) and anterior choroidal artery (AChA). Importantly, the vascular supply pattern varies between individuals (and hemispheres). While hippocampal blood supply usually depends on the PCA, the AChA additionally contributes in ca. 45% of hemispheres (“augmented supply”) and primarily supplies the hippocampal head. By means of high-resolution 7T time-of-flight MR angiography (ToF MRA), Spallazzi et al., 2018 were able to visualize the small cerebral arteries in the medial temporal lobe in vivo. Perosa et al. (2020, Brain) utilized this MRI sequence in older adults with and without cerebral small vessel disease (CSVD) and discovered that an augmented hippocampal vascular supply by the PCA and AChA in at least one hemisphere was linked to better verbal memory and better global cognition.
In a follow-up study, our group demonstrated that larger gray matter volumes in individuals with an augmented vascular supply of the hippocampus were specifically observed in the anterior MTL, including the anterior hippocampus and entorhinal cortex (Vockert et al., 2021, Brain Comm). However, our data did not reveal any evidence that an augmented vascularization conveys resistance or resilience against vascular pathology itself. Overall this suggests that an augmented hippocampal vascularization might contribute to brain maintenance or reserve (higher or preserved brain volume) and might help individuals with vascular disease to better cope with vascular pathology. Future studies are needed to support this hypothesis.
In our previous work, vessel classifications were based on the qualitative observation (visual rating) of specific arteries nurturing the hippocampus, which has certain limitations. In collaboration with Dr. Hendrik Mattern, Prof. Oliver Speck and Prof. Stefanie Schreiber in Magdeburg, we recently applied a novel quantitative methodology for assessing vascularization in the hippocampus. Vessel Distance Mapping (VDM) constitutes a new approach for assessing vascularization as a continuous metric beyond binary classification, by performing data-driven analysis of vessel patterns and vessel density (e.g. Mattern et al., ESMRMB 2020, 2021). Based on a vessel segmentation, VDM computes for each voxel the distance to its closest vessel, providing per voxel an estimate of how close the surrounding vasculature is. In a recent study, in which we applied VDM on the hippocampus (Garcia-Garcia*, Mattern* et al, AAIC 2022) we found that lower values of VDM-metrics reflecting lower distances to vessels in the hippocampus were associated with better cognitive performance specifically in subjects with CSVD. These data suggest that cognitive resilience in the face of vascular disease might be mediated by the vessel pattern and vessel density in proximity to the hippocampus. In future studies, we will use VDM to confirm and extend these findings with respect to resilience towards AD pathology.
In our collaborative project number “B04” funded by the CRC1436 (2021-2024) that is co-lead by Prof. Stefanie Schreiber and Prof. Esther Kühn, we are now investigating if hippocampal vascularization patterns constitute a factor for resistance or resilience in relation to vascular and early tau pathology by means of multi-modal imaging, including 3T and 7T MRI: https://sfb1436.de/projects/effects-of-hippocampal-vascularization-patterns-on-the-neural-resources-of-mtl-neurocognitive-circuits/. Specifically, we study how hippocampal vascularization relates to regional blood flow, volume of MTL subregions, myelin and functional circuits in young adults and in cognitively unimpaired? older adults.
In addition, we will investigate how hippocampal vascularization patterns affect individual learning behavior, learning related activity and benefits from real-world navigation training (together with Dr. Nadine Diersch). To do so, we will use the “Explore”-app, developed by Dr. Nadine Diersch to track real-world navigation behavior (Marquardt et al., 2022; AAIC 2022). We will examine the effect of the hippocampal vascularization on hippocampal activity when learning a spatial environment and how activity changes after a 3-week intervention using the “Explore”-app to train spatial navigation abilities in older adults in multiple sessions on the local medical campus.
Several concepts have been used to account for “resilience” against age- and disease-related changes including brain reserve, cognitive reserve, and brain maintenance. Cognitive reserve (CR) to a better than expected cognitive performance given the degree of age-related brain changes or disease (Stern et al., 2018). There are many potential mechanisms implicated in the complex construct of cognitive reserve, that likely rely on both structural and functional brain mechanisms. FMRI allows for the characterization and measurement of functional brain processes, thus offering the unique opportunity to capture the neural implementation of cognitive reserve. More specifically, individual differences in brain function or patterns of brain activity during fMRI tasks may explain the differential susceptibility to pathological burden.
A recent consensus definition (https://reserveandresilience.com/framework/) provides an operational framework to study cognitive reserve. In a collaboration with Gabriel Ziegler we employed this framework in a sophisticated approach including dimensionality reduction techniques and a multivariate moderation model in order to identify an fMRI-based activation pattern whose expression is related to cognitive reserve in AD (Vockert et al; AAIC 2022). The approach additionally enables the calculation of a cognitive reserve score, which could be useful in other contexts like the characterization of individuals with high levels of CR. While there is also an extended interest in task-invariant, generic components of CR, our focus so far has been on the specific context of successful memory encoding, which is of particular clinical significance because of its relevance in aging and dementia. The ultimate goal would be to leverage longitudinal studies in order to assess the utility of CR scores in predicting cognitive decline, i.e. the utility of CR networks in terms of their ability to modulate trajectories of cognitive decline. Longitudinal approaches could further enable us to elucidate the degree to which cognitive reserve is dynamic over time and amenable to interventions.