Potential CSF Biomarkers Detect Roused Microglia in AD, FTD

As immune sentinels of the brain, microglia perk up during the preclinical stages of Alzheimer’s and related diseases. How they respond plays a strong hand in how disease progresses. As scientists ramp up development of therapies aimed at microglia, they need sensitive biomarkers to gauge microglial activation. Scientists want multiple markers, as they have come to appreciate that microglia are nimble and quite varied in how they change states in response to pathologic aggregates.

  • Proteomes of TREM2 versus GRN knockout mice and cells suggest markers of microglial activation.
  • A panel of six CSF proteins distinguish people with FTD-GRN or AD from controls.
  • Three—FABP3, GDI1, and MDH1—signal Aβ deposition among people with MCI.

To that end, researchers led by Christian Haass of the German Institute for Neurodegenerative Diseases in Munich report the discovery of a panel of proteins that rise in CSF as microglia respond to proteopathy brewing in the brain. Their approach: probing the proteomes of knockout mice and human microglial cells on opposing ends of the microglial activation spectrum. The researchers identified six proteins that distinguish people with frontotemporal dementia caused by a progranulin mutation (FTD-GRN) from controls. What’s more, three of the proteins—FABP3, GDI1, and MDH1—also worked as potential AD biomarkers, correlating with amyloid deposition among people with mild cognitive impairment. The findings were posted June 18 on BioRxiv.

These proteins could provide information during clinical trials for drugs targeting microglia. They might also aid in prognosis for people with neurodegenerative diseases, or any other brain disorder influenced by microglia, said co-author Henrik Zetterberg of the University of Gothenburg in Sweden.

To zero in on markers of microglial activation, first author Ida Pesämaa and colleagues started by comparing the microglial proteomes of mice devoid of the microglial receptor TREM2 with those missing the lysosomal protein progranulin (GRN). Why these two models? Essentially, they enlist microglia on opposite ends of their activation spectrum. Whereas microglia lacking TREM2 remain locked in a homeostatic state, those missing progranulin become hyperactivated (Götzl et al., 2019). Mirroring this dichotomy, microglia from GRN knockout mice were brimming with proteins previously implicated in the disease-associated microglia (DAM) signature, including ApoE, Lgals3, Lyz2, and Clec7a. Some of these were also found in the mice’s CSF. Conversely, microglia in TREM2 knockouts expressed a glut of homeostatic proteins.

The researchers next turned to human cells. They compared the proteomes of human induced pluripotent stem cell (iPSC)-derived microglia (hiMGL) lacking either TREM2 or GRN. While lysates fromTREM2 knockout microglia contained more homeostatic proteins than wild-type cells, GRN knockout microglia had cranked up expression of a bevy of activation proteins, nearly half of which had also been elevated in microglia from GRN knockout mice. A subset of these proteins also rose within the medium of the cultured cells: Might they be found in human CSF?

To investigate this, Pesämaa and colleagues compared the CSF proteomes of 11 people with FTD-GRN versus 12 controls. Zetterberg said he was positively surprised by what they found: Of the 62 proteins that were elevated in FTD-GRN CSF, 26 had also been cranked up by cultured GRN knockout hiMGLs. This cast these 26 proteins as reflective of the heightened activation state of microglia in the diseased brain. It also supported the idea that disease-related changes in microglia—a relatively sparse type of brain cell—can be detected in CSF.

From these 26 proteins, the researchers selected six of the most abundantly upregulated ones to validate further. The so-called “panel 6” comprised fatty acid binding protein 3 (FABP3), malate dehydrogenase 1(MDH1), GDP dissociation inhibitor-1 (GDI1), macrophages capping protein (CAPG), CD44, and glycoprotein NMB (GPNMB).

Microglial Activation Panel. All six proteins in microglial activation panel were up in both FTD-GRN CSF and in media from GRN KO hiMGLs. Some were also elevated in hiMGL lysate, or in GRN KO mouse microglia or CSF. Circle size indicates statistical significance; color indicates upregulation relative to wild-type. [Courtesy of Pesämaa et al., bioRxiv, 2023.]

Pesämaa put the diagnostic potential of the panel to a first test. When combined, the panel 6 proteins separated FTD-GRN from controls with an area under the ROC curve of 0.87—in other words, with an accuracy of 87 percent. FABP3 alone performed even better, sorting cases from controls in the present cohort with 93 percent accuracy.

How do these microglial activation markers behave in people with AD? To find out, the scientists measured panel 6 proteins in CSF of 478 participants in the EMIF-AD cohort, which included 126 cognitively normal controls, 61 with subjective cognitive impairment, 198 with MCI, and 93 with AD. Three of the proteins—FABP3, MDH1, and GDI1—were significantly elevated in people MCI and AD relative to the other two groups. Notably, these same three proteins tracked with brain amyloid: Among people diagnosed with MCI, they were more abundant in people with evidence of amyloid. Strikingly, the panel 6 proteins distinguished amyloid-positive people with AD from amyloid-negative healthy controls with an accuracy of 94 percent, while a combination of FABP3, GDI1, and MDH1 did so with 92 percent accuracy.

Together, the findings suggest that markers of microglial responses can be detected in the CSF of people with neurodegenerative disease. In the case of FTD-GRN, microglia deficient in progranulin are contending with a build-up of unprocessed lysosomal substrates. In people with MCI and AD, the cells must deal with Aβ aggregates, tau tangles, and struggling neurons. Either way, microglia are under stress, Pesämaa said.

Collectively, the panel 6 markers converge on lipid metabolism as a major feature of the microglial response in these stressful situations. For example, FABP3 is involved in uptake and processing of long-chain fatty acids and, along with ApoE, has been proposed as a marker for lipid metabolism in AD (Furuhashi and Hotamisligil, 2008; Dulewicz et al., 2021). GPNMB is reportedly elevated among microglia surrounding plaques, especially among cells loaded with lipids (Hüttenrauch et al., 2018). Finally, GPNMB is a ligand for the CD44 receptor, which itself has been implicated in lipid metabolism (Neal et al., 2018). 

Zetterberg believes the results may apply even more broadly than neurodegenerative disease, perhaps picking up microglial rumblings in other disorders such as COVID-provoked brain fog. To find out, the researchers are testing the panel in banked CSF samples from different cohorts, including longitudinal ones. They hope to understand more broadly how these markers change as diseases progress. Ultimately, the markers could prove useful in clinical trials aimed at microglia, or even in the context of amyloid-targeted therapies, Zetterberg said.—Jessica Shugart

Comments

  1. As we enter the age of plausible microglia-targeted therapeutics, it is essential that relevant biomarkers, applicable across neurodegenerative diseases, be identified.

    Positron emission tomography (PET) studies with tracers targeting the TSPO protein have characterized neuroinflammation as a common feature across these diseases, and as a key player in defining prognosis and disease progression in both frontotemporal dementia and Alzheimer’s disease (Malpetti et al., 2020; Pascoal et al., 2021; Malpetti et al., 2023). However, TSPO PET currently cannot distinguish microglial phenotypes and related processes. To this end, complementary in vivo biomarkers are needed to stratify patients, to ascertain target engagement, and to monitor effects of microglia-modulating therapies that are in the development pipeline, as neuroinflammation has become a principal candidate target of AD drug development (Cummings et al., 2023).

    The effort to find such markers requires a deeply translational approach. This innovative study by Pesämaa, Haass, and colleagues provides a great example of how one can tackle biomarker development for this field and fill knowledge gaps by combining preclinical (murine and human IPSC-derived microglia) and clinical data (biospecimens from people living with dementia) in a translational workflow.

    Using this approach, the authors identify a panel of six proteins that were upregulated in the CSF of FTD progranulin mutation carriers. A subset of three of these proteins was also elevated in people with AD and mild cognitive impairment (thus able to detect early changes) and sensitive to participants’ amyloid status.

    This opens up the possibility that proteomic approaches such as this one may, when combined with complementary markers including PET, be able to focus in on microglial activation-related processes.

    Future studies will need to validate these markers across different disorders and assess whether levels of these proteins co-vary with and/or predict clinical outcome/severity. It will also be key to determine whether the CSF-based findings may translate to more readily available and scalable blood markers.

    Pesämaa, Haass, and colleagues are to be congratulated for taking a further crucial step toward novel, clinically plausible cross-diagnostic and translational microglial biomarkers.

    References:

    Cummings J, Lee G, Zhong K, Fonseca J, Taghva K. Alzheimer's disease drug development pipeline: 2021. Alzheimers Dement (N Y). 2021;7(1):e12179. Epub 2021 May 25 PubMed.

    Malpetti M, Kievit RA, Passamonti L, Jones PS, Tsvetanov KA, Rittman T, Mak E, Nicastro N, Bevan-Jones WR, Su L, Hong YT, Fryer TD, Aigbirhio FI, O'Brien JT, Rowe JB. Microglial activation and tau burden predict cognitive decline in Alzheimer's disease. Brain. 2020 May 1;143(5):1588-1602. PubMed.

    Malpetti M, Cope TE, Street D, Jones PS, Hezemans FH, Mak E, Tsvetanov KA, Rittman T, Bevan-Jones WR, Patterson K, Passamonti L, Fryer TD, Hong YT, Aigbirhio FI, O'Brien JT, Rowe JB. Microglial activation in the frontal cortex predicts cognitive decline in frontotemporal dementia. Brain. 2023 Aug 1;146(8):3221-3231. PubMed.

    Pascoal TA, Benedet AL, Ashton NJ, Kang MS, Therriault J, Chamoun M, Savard M, Lussier FZ, Tissot C, Karikari TK, Ottoy J, Mathotaarachchi S, Stevenson J, Massarweh G, Schöll M, de Leon MJ, Soucy JP, Edison P, Blennow K, Zetterberg H, Gauthier S, Rosa-Neto P. Microglial activation and tau propagate jointly across Braak stages. Nat Med. 2021 Sep;27(9):1592-1599. Epub 2021 Aug 26 PubMed. Correction.

    View all comments by Maura Malpetti
    • User Profile Image Marc Suárez-Calvet
      BarcelonaBeta Brain Research Center; Hospital del Mar - Barcelona
    • Posted:

    Amid development of disease-modifying treatments for Alzheimer's disease, including emerging microglia-modulating therapies, we need to develop reliable biomarkers for evaluating microglial dynamics in humans. CSF sTREM2 has been extensively studied as a marker of the TREM2-mediated microglial response in AD and, in most studies, higher CSF sTREM2 levels have been associated with a more favorable prognosis for patients (Morenas-Rodríguez et al., 2022; Ewers et al., 2019; Suárez-Calvet et al., 2016; Pereira et al., 2022). 

    However, the microglial response to the disease is intricate and dynamic, involving various cell states. We must identify additional markers beyond sTREM2 that better reflect the diversity of the microglial response and its potential as a target for therapeutic modulation.

    In their study, Pesämaa et al. took an innovative approach by investigating two opposing microglial phenotypes. They analyzed the proteome of microglia and CSF from genetically modified mouse models lacking TREM2 (locked in a homeostatic state) and GRN (hyperactivated microglia) (Götzl et al., 2019). Additionally, they examined iPSC-derived microglia.

    The authors successfully identified specific proteins associated with opposite microglial activation states. These findings were further validated by analysing CSF from patients with frontotemporal dementia carrying GRN mutations. The authors proposed a panel of six proteins as potential indicators for microglial reactivity. Notably, three of these (FABP3, GDI1, MDH1) were capable of differentiating amyloid-positive from amyloid-negative cases with mild cognitive impairment. It is worth mentioning that some of these proteins are also implicated in lipid dysmetabolism, as genetic evidence has linked several Alzheimer's disease-associated risk loci to lipid and microglial metabolism.

    Microglial state definitions predominantly rely on transcriptomic profiling. While numerous studies have explored single glial reactivity markers in human fluids, the question arises as to whether we can establish a marker profile that accurately reflects microglial state dynamics. This study, along with a recent investigation by the BioFINDER group—which investigates a set of markers of the disease-associated microglial activation stage 2 (Pereira et al, 2022)—is paving the way for future research in glial biomarkers.

    Moving forward, it is crucial to further investigate and validate these candidate microglial biomarkers in larger, longitudinal cohorts covering the entire Alzheimer's disease continuum, as well as other neurodegenerative cohorts. Understanding the microglial states associated with better or worse prognoses could aid in the design of more effective microglial-modulating therapies. Additionally, these microglial biomarkers may hold clinical utility in monitoring therapeutic response and identifying patients who would benefit most from novel treatments targeting microglial reactivity, but also amyloid deposition.

    It is conceivable that the individual response to anti-amyloid drugs may vary depending on the predominant microglial state response exhibited by each individual.

    References:

    Ewers M, Franzmeier N, Suárez-Calvet M, Morenas-Rodriguez E, Caballero MA, Kleinberger G, Piccio L, Cruchaga C, Deming Y, Dichgans M, Trojanowski JQ, Shaw LM, Weiner MW, Haass C, Alzheimer’s Disease Neuroimaging Initiative. Increased soluble TREM2 in cerebrospinal fluid is associated with reduced cognitive and clinical decline in Alzheimer's disease. Sci Transl Med. 2019 Aug 28;11(507) PubMed.

    Götzl JK, Brendel M, Werner G, Parhizkar S, Sebastian Monasor L, Kleinberger G, Colombo AV, Deussing M, Wagner M, Winkelmann J, Diehl-Schmid J, Levin J, Fellerer K, Reifschneider A, Bultmann S, Bartenstein P, Rominger A, Tahirovic S, Smith ST, Madore C, Butovsky O, Capell A, Haass C. Opposite microglial activation stages upon loss of PGRN or TREM2 result in reduced cerebral glucose metabolism. EMBO Mol Med. 2019 Jun;11(6) PubMed.

    Morenas-Rodríguez E, Li Y, Nuscher B, Franzmeier N, Xiong C, Suárez-Calvet M, Fagan AM, Schultz S, Gordon BA, Benzinger TL, Hassenstab J, McDade E, Feederle R, Karch CM, Schlepckow K, Morris JC, Kleinberger G, Nellgard B, Vöglein J, Blennow K, Zetterberg H, Ewers M, Jucker M, Levin J, Bateman RJ, Haass C, Dominantly Inherited Alzheimer Network. Soluble TREM2 in CSF and its association with other biomarkers and cognition in autosomal-dominant Alzheimer's disease: a longitudinal observational study. Lancet Neurol. 2022 Apr;21(4):329-341. PubMed.

    Pereira JB, Janelidze S, Strandberg O, Whelan CD, Zetterberg H, Blennow K, Palmqvist S, Stomrud E, Mattsson-Carlgren N, Hansson O. Microglial activation protects against accumulation of tau aggregates in nondemented individuals with underlying Alzheimer's disease pathology. Nat Aging. 2022 Dec;2(12):1138-1144. Epub 2022 Nov 28 PubMed.

    Suárez-Calvet M, Kleinberger G, Araque Caballero MÁ, Brendel M, Rominger A, Alcolea D, Fortea J, Lleó A, Blesa R, Gispert JD, Sánchez-Valle R, Antonell A, Rami L, Molinuevo JL, Brosseron F, Traschütz A, Heneka MT, Struyfs H, Engelborghs S, Sleegers K, Van Broeckhoven C, Zetterberg H, Nellgård B, Blennow K, Crispin A, Ewers M, Haass C. sTREM2 cerebrospinal fluid levels are a potential biomarker for microglia activity in early-stage Alzheimer's disease and associate with neuronal injury markers. EMBO Mol Med. 2016 May 2;8(5):466-76. PubMed.

    View all comments by Marc Suárez-Calvet

  3. There is an urgent need to identify biomarkers associated with microglial activation in order to increase its protective mechanisms against pathology. In this study, the authors observed that GRN KO mice, which show microglial hyperactivation, present upregulation of a panel of six markers. These markers were increased in frontotemporal dementia cases and three of them—FABP3, MDH1, and GDI1—were also elevated in AD patients in response to early amyloid accumulation.

    These findings shed new light into biomarkers that can inform us about specific microglial states, which can increase their protective effects in clinical trials. Replication in other cohorts is now needed to confirm these interesting findings.

    View all comments by Joana Pereira

  4. Here, Pesämaa et al. used two different models—GRN KO mice with TREM2 KO mice, and hiMGL with GRN mutations—to identify a panel of six biomarkers for microglia activation. They then validated the upregulation of some of these biomarkers in both FTD and AD CSF. It is nice to see that markers identified in mouse models and in vitro can be validated in human patients.

    Among the six biomarkers identified, GPNMB and FABP3 had been previously identified as DAM markers in the 5XFAD model (Keren-Shaul, 2017), whereas CD44 was identified as a reactive astrocyte marker (Qian et al., 2023). hus, as the authors mention in their manuscript, the identified biomarkers are not microglia-specific, but more likely microglia activity-dependent.

    In the next steps, it would be important to correlate the biomarker levels with disease progression and pathological changes, for example, to correlate the biomarker levels with amyloid-PET or tau-PET in AD, so that the biomarkers can be used to predict disease progression and to speculate on the time window for targeting microglia therapeutically.

    References:

    Keren-Shaul H, Spinrad A, Weiner A, Matcovitch-Natan O, Dvir-Szternfeld R, Ulland TK, David E, Baruch K, Lara-Astaiso D, Toth B, Itzkovitz S, Colonna M, Schwartz M, Amit I. A Unique Microglia Type Associated with Restricting Development of Alzheimer's Disease. Cell. 2017 Jun 15;169(7):1276-1290.e17. Epub 2017 Jun 8 PubMed.

    Qian K, Jiang X, Liu ZQ, Zhang J, Fu P, Su Y, Brazhe NA, Liu D, Zhu LQ. Revisiting the critical roles of reactive astrocytes in neurodegeneration. Mol Psychiatry. 2023 Apr 10; PubMed.

    View all comments by Yingyue Zhou

Make a Comment

To make a comment you must login or register.

References

Paper Citations

  1. Götzl JK, Brendel M, Werner G, Parhizkar S, Sebastian Monasor L, Kleinberger G, Colombo AV, Deussing M, Wagner M, Winkelmann J, Diehl-Schmid J, Levin J, Fellerer K, Reifschneider A, Bultmann S, Bartenstein P, Rominger A, Tahirovic S, Smith ST, Madore C, Butovsky O, Capell A, Haass C. Opposite microglial activation stages upon loss of PGRN or TREM2 result in reduced cerebral glucose metabolism. EMBO Mol Med. 2019 Jun;11(6) PubMed.
  2. Furuhashi M, Hotamisligil GS. Fatty acid-binding proteins: role in metabolic diseases and potential as drug targets. Nat Rev Drug Discov. 2008 Jun;7(6):489-503. PubMed.
  3. Dulewicz M, Kulczyńska-Przybik A, Słowik A, Borawska R, Mroczko B. Fatty Acid Binding Protein 3 (FABP3) and Apolipoprotein E4 (ApoE4) as Lipid Metabolism-Related Biomarkers of Alzheimer's Disease. J Clin Med. 2021 Jul 6;10(14) PubMed.
  4. Hüttenrauch M, Ogorek I, Klafki H, Otto M, Stadelmann C, Weggen S, Wiltfang J, Wirths O. Glycoprotein NMB: a novel Alzheimer's disease associated marker expressed in a subset of activated microglia. Acta Neuropathol Commun. 2018 Oct 19;6(1):108. PubMed.
  5. Neal ML, Boyle AM, Budge KM, Safadi FF, Richardson JR. The glycoprotein GPNMB attenuates astrocyte inflammatory responses through the CD44 receptor. J Neuroinflammation. 2018 Mar 8;15(1):73. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. Pesämaa I, Müller SA, Robinson S, Darcher A, Paquet D, Zetterberg H, Lichtenthaler SF, Haass C. A MICROGLIAL ACTIVITY STATE BIOMARKER PANEL DIFFERENTIATES FTD-GRANULIN AND ALZHEIMER'S DISEASE PATIENTS FROM CONTROLS. bioRxiv. 2023 Jun 16; PubMed.

Annotate

To make an annotation you must Login or Register.