. Integrative in situ mapping of single-cell transcriptional states and tissue histopathology in a mouse model of Alzheimer's disease. Nat Neurosci. 2023 Mar;26(3):430-446. Epub 2023 Feb 2 PubMed.

Recommends

Please login to recommend the paper.

Comments

  1. This is a beautiful paper using spatial transcriptomics to depict the cellular phase of Alzheimer pathology (De Strooper and Karran, 2016). The technology brings us a huge step forward from our transcriptomic paper in 2020 (Chen et al., 2020). With the technology available to us at that time, we had a resolution of 100 µm allowing us to analyze small tissue domains. We could not perform protein staining on the same section and therefore used two adjacent slides to locate the amyloid plaques.

    Zheng et al. provide real single-cell resolution combined with protein staining of amyloid plaques and tau pathology in situ. We feel very pleased to see that some of the major findings of our work hold upon this deeper scrutiny. The PIGs (plaque-induced genes) we described depict the microglia and astroglia responses around the plaques and their coordinated reaction to these pathologies, as seen here with disease-associated microglia and astrocytes. We also noticed in our paper the oligodendrocyte reaction contributing to the overall cell response, something that was a big surprise for us at that time because not many researchers were thinking about oligodendrocytes playing a role in the disease process.

    It is clear that with this type of analysis, the field is transitioning into a new phase of understanding of the pathogenesis of Alzheimer’s disease. The emerging picture is that the amyloid deposits in the brains of mice and humans are not innocent. They induce a mainly microglia-driven multicellular response, most similar to a chronic inflammation. The fact that most of the risk genes for Alzheimer’s disease are expressed in microglia suggests that this response is central to the disease progression (Sierksma et al., 2020Sierksma et al., 2020). It is not the amyloid plaque that causes the dementia, but the cellular response to these lesions that determines whether somebody becomes clinically sick or not.  

    References:

    . The Cellular Phase of Alzheimer's Disease. Cell. 2016 Feb 11;164(4):603-15. PubMed.

    . Spatial Transcriptomics and In Situ Sequencing to Study Alzheimer's Disease. Cell. 2020 Aug 20;182(4):976-991.e19. Epub 2020 Jul 22 PubMed.

    . Novel Alzheimer risk genes determine the microglia response to amyloid-β but not to TAU pathology. EMBO Mol Med. 2020 Mar 6;12(3):e10606. Epub 2020 Jan 17 PubMed.

    . Translating genetic risk of Alzheimer's disease into mechanistic insight and drug targets. Science. 2020 Oct 2;370(6512):61-66. PubMed.

    View all comments by Bart De Strooper
  2. Using in-house-developed STARmap PLUS for single-cell spatial transcriptomics, Zeng et al. established a cellular map that reflects disease pathology in the TauPS2APP transgenic model of Alzheimer’s disease. The beautiful data shown here nicely corroborated previous identifications of disease-associated microglia and astrocytes, and comprehensively provided a spatial relationship of these disease-associated populations with plaque and tau pathologies. The results showed that the STARmap PLUS technology can be widely used to study changes in gene signatures in 4D, combining space and time, through disease progression.

    Compared to mouse models, spatial transcriptomic information from human patients is far sparser. STARmap PLUS will greatly aid in our understanding of disease progression when applied to postmortem human samples. For example, whether the core-shell structures are similarly presented in human tissues is unclear.

    We and others have previously identified a reactive oligodendrocyte population in the 5XFAD mice and various CNS pathologies (Zhou et al., 2020Shen et al., 2021; Kenigsbuch et al., 2022). This population was not highlighted in the current study, likely because one of the marker genes for this oligodendrocyte population, Serpina3n, was not included in the target gene list. Nonetheless, the Oligo2 cluster enriched in TauPS2APP might be the corresponding reactive oligodendrocytes. It is interesting to see that the Oligo2 cluster is enriched around p-tau but not as much around the amyloid plaques, suggesting differential effects of p-tau and amyloid on oligodendrocyte responses.

    References:

    . Human and mouse single-nucleus transcriptomics reveal TREM2-dependent and TREM2-independent cellular responses in Alzheimer's disease. Nat Med. 2020 Jan;26(1):131-142. Epub 2020 Jan 13 PubMed. Correction.

    . Multiple sclerosis risk gene Mertk is required for microglial activation and subsequent remyelination. Cell Rep. 2021 Mar 9;34(10):108835. PubMed.

    . A shared disease-associated oligodendrocyte signature among multiple CNS pathologies. Nat Neurosci. 2022 Jul;25(7):876-886. Epub 2022 Jun 27 PubMed.

    View all comments by Yingyue Zhou
  3. The authors present an improved spatial mapping method termed STARmap PLUS, which is able to capture both cell-specific transcriptional states and pathological hallmarks of Alzheimer’s disease (AD), such as Aβ plaques and neurofibrilarly tangles of tau. Using STARmap PLUS on the TauPS2APP mouse model (which exhibits both plaques and tangles), they highlight the enrichment of neurodegenerative microglia (MGnD or DAM) in immediate proximity of plaques, while disease-associated astrocytes (DAA) are frequent within the secondary perimeter. A disease-associated subtype of oligodendrocytes was found predominatly around tau-positive excitatory neurons in the CA1 region; however, this was predominantly at later stages (13 months compared to 8 months).

    The study provides an important advance to our ability of mapping disease pathology with functional and identifiable cell states. STARmap PLUS outperforms current spatial expression methods, and reproduced key cell substates such as MGnD and DAA. The current limitation is the curated selection of 2,766 genes and the capture of about 70k cells. The technical constraint might explain why the recently identified interferon responsive microglia (IRM) subtype was not captured (Sala Frigerioet al., 2019; Dorman et al., 2022; Kaya et al., 2022). 

    Due to the two-dimesional restriction of the method, the spatial alterations to endothelial cells were also not captured. The importance of blood-brain barrier leakage, amyloid deposition in blood vessels, and cerebral amyloid angiopathy have been reported as major factors in AD (Kisler et al., 2017; Ellis et al., 1996; Johnson et al., 2007). Hence, associating particular changes in endothelial cells to local levels of amyloid deposition is of major interest to the the field. Furthermore, the technology holds the potential to highlight localized disease crosstalk, for example between MGnD and DAA around plaques.

    Overall, STARmap PLUS shifts our understanding of individual cell substates themselves toward their simultanous pathological location, and future pertubation approaches will hopefully allow the identification of disease-promoting or -responsive mechanisms.

    References:

    . The Major Risk Factors for Alzheimer's Disease: Age, Sex, and Genes Modulate the Microglia Response to Aβ Plaques. Cell Rep. 2019 Apr 23;27(4):1293-1306.e6. PubMed.

    . A type I interferon response defines a conserved microglial state required for effective neuronal phagocytosis. bioRxiv. 2022 Feb 22; PubMed.

    . CD8+ T cells induce interferon-responsive oligodendrocytes and microglia in white matter aging. Nat Neurosci. 2022 Nov;25(11):1446-1457. Epub 2022 Oct 24 PubMed.

    . Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer disease. Nat Rev Neurosci. 2017 Jul;18(7):419-434. Epub 2017 May 18 PubMed.

    . Cerebral amyloid angiopathy in the brains of patients with Alzheimer's disease: the CERAD experience, Part XV. Neurology. 1996 Jun;46(6):1592-6. PubMed.

    . Imaging of amyloid burden and distribution in cerebral amyloid angiopathy. Ann Neurol. 2007 Sep;62(3):229-34. PubMed.

    View all comments by Oleg Butovsky
  4. Some technical thoughts: High-throughput in situ methods, such as STARmap and merscope, are going to continue to afford a greater degree of confidence in spatial transcriptomics methods moving forward. The ability to have individual cell resolution is a real bonus, and subcellular resolution is going to become the norm.

    This is something that spotted oligo array systems, e.g. Visium, are unable to provide currently, though promised updates with VisiumHD may overcome some of these concerns with improved capture areas and greatly increased resolution. As a side note, however, if the fiducial area provides too many spots, we will be unable to afford the sequencing on these spots, as a predicted 1 million spots will be a huge cost barrier using current sequencing technology. Maybe Ultima or NOVAX from Illumina will fix this.

    One remaining concern to keep in mind is that, while Visium provides lower fidelity, it does provide unbiased gene analysis without a requirement for a priori knowledge of “what to look for.” In situ based methods such as merscope and STARmap, while providing much higher resolution, require the selection of known marker genes to put into your panel of probes. This is less of a problem if people are completing paired sc/snRNAseq in experiments where they also take sections for spatial transcriptomics. I would also suggest doing this for Visium, so that you can make cell-type/sub-state specific modules of genes to probe back into the data.

    View all comments by Shane Liddelow
  5. We are happy to see that an amyloid-induced, multicellular, gene co-regulation network—i.e., 57 plaque-induced genes or PIGs that we identified in APPNL-G-F mice—has been cross-validated in multiple AD mouse models including 5XFAD, APPPS1, and now in TauPS2APP mice by Zeng and colleagues (Zhou et al., 2020; Kenigsbuch et al., 2022; Castranio et al., 2022). The results consistently show an activated glial niche surrounding amyloid plaques with APOE+/C1Q+ microglia as the core and GFAP+/C3+ astrocyte and SERPINA3+/C4B+/H2-D1+/B2m+ oligodendrocyte as the shell.

    PIGs have been identified via different technologies and analyses. We did so via spatial transcriptomics validated by in situ sequencing (Chen et al., 2020). We performed spatial transcriptomics, i.e., 2D RNA-Seq array, to unbiasedly cluster genes with similar spatial expression pattern from a pool of more than 15,000 genes across more than 10,000 microenvironments at 100 micrometer resolution. Protein staining on tissue sections adjacent the spatial transcriptomics section indicated one network, i.e. PIGs, co-localized with amyloid plaque, and the strength of gene co-expression gradually increased with Aβ level. Therefore, these 57 PIGs are not only enriched around plaques, as identified in many papers, but also co-expressed across diverse microenvironments of brain regions, ages, and genotypes with varying degrees of amyloid stress. This indicates that PIGs function together as a multicellular mechanism against amyloid stress, relationships that could not be recognized using single-nuclei RNA-Seq or in situ technologies with low throughput and sensitivity.

    We ran in situ sequencing to detect the 57 PIGs and many cell type markers, and to locate amyloid plaques via immunostaining on the same tissue section used for RNA-Seq, as did Zeng and colleagues. We thus validated enrichment of 57 PIGs around amyloid plaques at microscopic resolution in mouse brains and partially in human brains. However, due to the low sensitivity of in situ technology available at that time, we were unable to analyze gene expression at pseudo-cell resolution, or its distance to plaques. This paper nicely increases the throughput from 100 to 2,766 genes with good sensitivity, allowing pseudo-cell analysis and proximity to pathology. It’s a beautiful study to visualize the cellular response to amyloid plaques with single-cell and spatial resolution.

    There is a trade-off between throughput, sensitivity, specificity, and resolution of spatial technologies. It’s exciting to see more and more technologies increasing throughput and sensitivity at microscope resolution to demultiplex RNA detection and become commercialized (He et al., 2021; Janesick et al., 2022; Chen et al., 2022). 

    Mouse models have contributed enormously to our understanding of AD; however, the utility of these mice as robust preclinical models has been challenged. Spatial omic technologies open the possibility to learn how cells function together against pathogenic hallmarks directly in human brains, and thus offer a fantastic platform to screen therapeutic targets of neurodegenerative diseases.

    References:

    . Human and mouse single-nucleus transcriptomics reveal TREM2-dependent and TREM2-independent cellular responses in Alzheimer's disease. Nat Med. 2020 Jan;26(1):131-142. Epub 2020 Jan 13 PubMed. Correction.

    . A shared disease-associated oligodendrocyte signature among multiple CNS pathologies. Nat Neurosci. 2022 Jul;25(7):876-886. Epub 2022 Jun 27 PubMed.

    . Microglial INPP5D limits plaque formation and glial reactivity in the PSAPP mouse model of Alzheimer's disease. Alzheimers Dement. 2022 Nov 30; PubMed.

    . Spatial Transcriptomics and In Situ Sequencing to Study Alzheimer's Disease. Cell. 2020 Aug 20;182(4):976-991.e19. Epub 2020 Jul 22 PubMed.

    . High-plex Multiomic Analysis in FFPE at Subcellular Level by Spatial Molecular Imaging. bioRxiv. July 19, 2022 bioRxiv

    . High resolution mapping of the breast cancer tumor microenvironment using integrated single cell, spatial and in situ analysis of FFPE tissue. bioRxiv. October 7, 2022 bioRxiv

    . Spatiotemporal transcriptomic atlas of mouse organogenesis using DNA nanoball-patterned arrays. Cell. 2022 May 12;185(10):1777-1792.e21. Epub 2022 May 4 PubMed.

    View all comments by Wei-Ting Chen

Make a Comment

To make a comment you must login or register.

This paper appears in the following:

News

  1. Higher-Resolution Spatial Transcriptomics Maps Mayhem Near Plaques