30 Jun 2022

The adage of astrocytes as doting neuronal nursemaids has been supplanted by the view that astrocytes play a muscular part in the function of the brain. Provoke them, and astrocytes can shift from supporter to destructor. Two new studies examined different reactive states astrocytes adopt in response to inflammatory stimuli, and which transcriptional regulators were pulling the strings.

Martin Kampmann of the University of California, San Francisco, and colleagues posted theirs on bioRxiv on June 3. Running CRISPR screens on cultured, induced astrocytes rattled by a cocktail of inflammatory cytokines, the scientists uncovered two flavors of reactive astrocyte that antagonized one another by churning out opposing cytokines. One was induced by STAT3 and pumped out IL-6, the other was inhibited by STAT3 and spewed out interferon. Evidence for these two astrocyte types, and more types, emerged in multiple datasets and situations, including AD brain. Another study, led by Michael Sofroniew of the University of California, Los Angeles, and published in Nature on May 25, combined genomic approaches to reveal a striking diversity of reactive stances astrocytes took in different inflammatory contexts.

If there is one take-home message from these complex papers, perhaps it is this: “What is really emerging here is the concept that astrocytes show a large number of gene-expression signatures in their reactive states,” Kampmann told Alzforum.

In recent years, scientists have started to see astrocytes mount dynamic reactions in response to different injuries and diseases of the brain. For example, a paper in 2017 reported that bacterial lipopolysaccharides (LPS) made astrocytes stop their neuronal support services and transform into a cantankerous, neurotoxic state dubbed “A1” (Jan 2017 news). Spurred by an inflammatory stew of microglial proteins—IL-1α, TNF-α, and C1q—these cells turned up in multiple situations, including the AD brain. Since then, studies have described more diversity in reactive astrocyte states that go well beyond this A1 state.

Which transcriptional regulators dictate the reactive path an astrocyte takes in response to a given inflammatory stimulus? This was the main question Kampmann's group addressed. First author Kun Leng and colleagues grew induced astrocytes (iAstrocytes) from human induced pluripotent stem cells, then exposed them to the IL-1α+TNFα+C1q stew. As had been reported for human and mouse astrocytes previously, iAstrocytes jolted in this way promptly stopped phagocytosing floundering synaptosomes or supporting the maturation of synapses, and instead spewed factors that harmed neurons. The inflammatory cocktail also evoked dramatic changes in the transcriptomes of the iAstrocytes. Notably, the gene encoding VCAM1—a cell adhesion sialoglycoprotein that helps usher peripheral immune cells into the CNS—was consistently upregulated by iAstrocytes in response to the inflammatory treatment.

Hunt for Regulators. CRISPR-competent astrocytes were infected with a library of small guide RNAs and treated with a cocktail of cytokines or vehicle. Searching for sgRNAs enriched in cells whose synaptosome phagocytosis (top) or VCAM1 expression were perturbed then pinpointed regulators of astrocyte reactivity. [Courtesy of Leng et al., bioRxiv, 2022.]

With this model in hand, the scientists wielded the power of CRISPR interference (CRISPRi) to screen for regulators of the iAstrocyte inflammatory response. They looked for genes that, when knocked down, changed the astrocytes' ability to phagocytose synaptosomes, or to turn up VCAM1. To complement the iCRISPR screen, the researchers used a bioinformatic approach called “master regulator analysis,” which integrated data across gene-expression datasets to predict which transcriptional regulatory networks were active in astrocytes.

Combining their loot from both these approaches, the scientists pulled out both expected and novel regulators of inflammatory reactivity in astrocytes. Unsurprisingly, they found a strong role for the NF-kB pathway, a known master inducer of inflammation. They also identified two distinct groups of genes that were important for different aspects of the inflammatory response. One involved transcription factors, such as STAT3, which are tied to the acute phase of systemic inflammation and induce IL-6 secretion. The other included genes involved in the interferon response.

Deploying a broad set of experimental approaches, including single-cell transcriptomics, targeted knockdown of specific genes, and tracing signaling pathways switched on by different cytokines, the researchers next uncovered two modes of astrocyte reactivity in response to the IL-1α+TNFα+C1q mix. Both depend upon the canonical activation of NF-kB. One was marked by activation of STAT3, which promoted expression of IL-6 as well as the complement protein C3; the other centered expression of interferons and VCAM1.

Importantly, while the triple inflammatory mix revved both types of astrocyte response, the two antagonized each other in an autocrine/paracrine fashion. In other words, IL-6 secreted by one type of astrocyte suppressed interferon production from the other, and vice versa. Furthermore, while IL-1α polarized astrocytes toward the STAT3/IL-6 type of response, TNFα skewed them toward the interferon reaction.

The researchers found evidence for both astrocyte states—which they called IL-1/IL-6 responsive and TNF/IFN responsive—in other published datasets, including in response to the same mix of cytokines, in response to LPS, and in mice with experimental acute encephalopathy (EAE).

In the AD brain, the researchers found an abundance of astrocytes expressing C3, particularly around amyloid plaques. These astrocytes tended not to express VCAM1, a marker of the TNF/IFN response pathway, supporting the idea that the two pathways antagonize each other.

To learn whether that these two astrocyte responses regulate each other in vivo, the scientists injected lipopolysaccharide (LPS) into the bellies of mice and measured what type of astrocyte response ensued in the brain by staining for markers of the IL-1/6-responsive and TNF/IFN-responsive pathways. Abdominal LPS induced both types of reactive astrocyte in the hippocampus; however, in mice lacking Stat3 in their astrocytes, the LPS response skewed toward the TNF/IFN-responsive state, casting STAT3 as a key regulator of astrocyte responses.

“We don’t know why some astrocytes end up being in one state versus another,” Leng told Alzforum. That said, once an astrocyte goes down a path, the cytokines it secretes likely further polarize that cell into that reactive state, while suppressing the opposing state in itself and its surrounding astrocytes, Leng said. That both types of astrocyte were detected across multiple disease contexts suggests that there is a balance between the two, which might tip differently across disorders.

More broadly, Kampmann emphasized that the IL-1/IL-6 and TNF/IFN-responsive pathways identified within the experimental conditions of this paper are two of many reactive signatures at play in the brain.

Sofroniew’s study laid out the case for a complex constellation of astrocyte reactive states. First author Joshua Burda and colleagues compared gene-expression changes in astrocytes in response to three different neurological insults in mice: permanent tissue damage after spinal cord injury, transient neuroinflammation triggered by acute systemic injection LPS, and chronic neuroinflammation triggered by EAE.

Of the 6,471 genes that were differentially expressed in response to at least one of the inflammatory insults, only 172 were shared by all three. Using differentially expressed genes to predict which upstream transcriptional regulators (TRs) orchestrated these responses, the researchers reported that 255—including transcription factors, chromatin modifiers, and protein- or RNA-binding proteins—controlled the lion’s share of the downstream transcriptional reactions. Twenty percent of these TRs were shared across all three insults; about half were unique to one.

How to draw meaning from such a plethora of data? The researchers parsed the data by mapping accessible chromatin by way of ATAC-Seq, immunohistochemistry, RNA sequencing, and conditional knockouts of specific TRs.

Their paper describes an architecture of astrocyte reactivity. As part of that, it reports that one transcriptional regulator can control different groups of genes in different disorders. For example, astrocyte conditional knockouts of the transcriptional regulators Stat3 and Smarca4 revealed that each controls different, minimally overlapping sets of genes in response to LPS injection versus spinal cord injury. Moreover, in some cases, the TRs were able to regulate a gene in different directions depending on the inflammatory context. For example, Smarca4 and Stat3 were each required to upregulate Slc14a2 and Rhof in response to LPS treatment, and to downregulate these two after a spinal cord injury.

Converging with Kampmann’s findings, Sofroniew’s study also found that knocking Stat3 out of astrocytes upped expression of interferon genes. This jibes with the idea that Stat3 promotes expression of genes that inhibit interferon expression, Leng told Alzforum. What’s more, knocking Stat3 out of astrocytes made mice sicklier following LPS injection; this could stem from too much interferon.

Burda and colleagues next sifted through published astrocyte gene-expression data from 15 diseases or injuries, including Alzheimer’s and Huntington’s, amyotrophic lateral sclerosis, mouse models of these diseases, as well as injury models.

Unsettlingly, perhaps, they found a profound divergence of astrocyte reactive signatures, with little overlap in the differentially expressed genes across the disorders. Nothing here, it appears, is simple. They did report some commonality at the level of transcriptional regulators, identifying a core group of 61 TRs responsible for sending astrocytes down their respective reactive paths across the disorders.

Together, the two papers offer a glimpse of a remarkable breadth of astrocyte reactive states in different contexts. Kampmann said the next step will be to move beyond gene-expression signatures and understand the functional consequences of different reactive states in disease.

“Overall, the message is that contrary to long-held beliefs, astrocyte responses to disorders are not homogenous and stereotypic and functionally passive, but instead are very diverse and context-specific,” Sofroniew wrote to Alzforum. “Last but not least, astrocyte responses can influence neurological outcomes in disorders.”—Jessica Shugart

Comments

1. • Hanseok Ko
     Johns Hopkins University
   • Posted: 30 Jun 2022

   Within the complex system of the brain, the non-neuronal cells do make a great impact on the origin of neurodegenerative diseases. Recent experimental reports have evidently proved that the reactive form of astrocyte activates via releasing factors such as IL-1α, TNF, and C1q secreted by active microglia, and results in the demise of neuronal and oligodendrocyte cells leading to the cause of neurodegenerative diseases such as PD, AD, and ALS. The reactive astrocyte loses its homeostatic functions such as neuron survival, synaptogenesis, phagocytosis, etc., while impacting neighboring supporting neuronal and non-neuronal cells (Liddelow et al., 2017). Further, blocking reactive astrocytes by prohibiting microglial activation via the GLP-1R agonist NLY01 showed neuroprotective effects in in vitro and in vivo models of PD and AD respectively (Yun et al., 2018; Park et al., 2021). Yet, it is still uncovered by which cellular mechanisms reactive astrocytes lose homeostatic functions and gain harmful outputs.

   The new findings here by the Kampmann lab have tried to identify cellular pathways controlling inflammatory astrocyte reactivity in a unique manner, by employing the latest and most robust technology of single-cell transcriptomics with CRISPR interference (CRISPRi) screening in human iPSC (hiPSC)-derived astrocytes. Currently, the methods being utilized in vitro and in vivo to understand astrocyte reactivity are not clearly established due to limitations such as isolation of primary astrocytes or long procedure of generation of hiPSC-derived astrocytes. Moreover, molecular profiling approaches such as RNA-Seq, which is presently being used to determine altered cellular mechanisms in inflammatory reactive astrocytes, are also limited as the methodology could not explain the causative pathways for controlling inflammatory reactivity.

   Within this study, the authors have developed a scalable method and utilized the latest application of CRISPR-based functional genomics (Kampmann, 2020). CRISPRi screens were combined with hiPSC screens to uncover the cellular pathways responsible for IL-1α, TNF, and C1q- induced inflammatory astrocyte reactivity within this study. Furthermore, the scalability and homogeneity of hiPSC-derived iAstrocytes were vitally important to performing pooled screens. The phenotypes of multiple hiPSC-derived iAstrocytes were maintained throughout the whole study. The authors performed the reactivity evaluation based on CRISPRi screens and the computational master regulator analysis (MRA) platform, where they preferentially considered the CRISPRi screens due to MRA application in astrocyte-diverse phenotypes.

   The results obtained from CRISPRi screens with single transcriptomics data on the influence of IL-1α+TNF+C1q are that they induce two distinct inflammatory reactive astrocyte signatures (IRAS) forms, i.e., IRAS1 and IRAS2, dependent upon canonical NF-κB driving through IL-6 and interferon signaling. The transcription factors (CEBP/D, NF-κB, and STAT3) at the upstream promote IRAS1 while inhibiting IRAS2 through IL-6. IFN-β acts through STAT1/2 and IRF1 to partially influence both IRAS1 and IRAS2. The obtained findings state that IL-1/IL-6 are IRAS1-responsive, while IRAS2 is referred to as TNF/IFN-responsive signaling.

   Furthermore, several studies are well-corroborated with this study, supporting its finding of STAT3’s role in inflammatory astrocyte reactivity (Herrmann et al., 2008; Kim et al., 2022; Ben Haim et al., 2015), while potential inhibition or deletion prevents the neurodegenerative disease, spinal cord injury, and its progression in vitro and in vivo (Reichenbach et al., 2019; Anderson et al., 2016). This new study also found upregulated vascular cell adhesion molecule 1(VCAM1) transcript-level expression in response to the IL-1α+TNF+C1q, and results in support of previous findings with a pro-inflammatory environment (Rubio et al., 2010; Labib et al., 2022). 

   Along with the inflammatory astrocyte reactivity cellular signaling in vitro data, Leng and co-workers performed a few in vivo mouse model experiments and also studied human AD and hypoxic-ischaemic encephalopathy, and obtained the overlapped inflammatory profile.

   This iAstrocytes platform is unique and will be greatly beneficial in future investigations using patient-derived hiPSCs in dissecting the impact of disease-associated mutations on inflammatory astrocyte reactivity. This appreciable work acts as an important reference to future approaches in characterizing and validating the practical outputs of IL-1/IL-6-responsive versus TNF/IFN-responsive astrocyte reactivity in animal models of neuroinflammation and neurodegeneration.

   The significant part of this advanced technical approach is an application that accelerates drug discovery. Furthermore, it can help in searching for novel cellular proteins involved in diverse cellular signaling using CRISPR pooled-screen with single-cell transcriptomics in various cell types such as microglia, oligodendrocytes, endoderm development, neuronal development, etc.

   Moreover, this will serve as a tool to characterize β-cell type function and mechanism and selectively monitor the development of novel therapeutics against several neurodegenerative diseases.

   References:

   Liddelow SA, Guttenplan KA, Clarke LE, Bennett FC, Bohlen CJ, Schirmer L, Bennett ML, Münch AE, Chung WS, Peterson TC, Wilton DK, Frouin A, Napier BA, Panicker N, Kumar M, Buckwalter MS, Rowitch DH, Dawson VL, Dawson TM, Stevens B, Barres BA. Neurotoxic reactive astrocytes are induced by activated microglia. Nature. 2017 Jan 26;541(7638):481-487. Epub 2017 Jan 18 PubMed.

   Yun SP, Kam TI, Panicker N, Kim S, Oh Y, Park JS, Kwon SH, Park YJ, Karuppagounder SS, Park H, Kim S, Oh N, Kim NA, Lee S, Brahmachari S, Mao X, Lee JH, Kumar M, An D, Kang SU, Lee Y, Lee KC, Na DH, Kim D, Lee SH, Roschke VV, Liddelow SA, Mari Z, Barres BA, Dawson VL, Lee S, Dawson TM, Ko HS. Block of A1 astrocyte conversion by microglia is neuroprotective in models of Parkinson's disease. Nat Med. 2018 Jul;24(7):931-938. Epub 2018 Jun 11 PubMed.

   Park JS, Kam TI, Lee S, Park H, Oh Y, Kwon SH, Song JJ, Kim D, Kim H, Jhaldiyal A, Na DH, Lee KC, Park EJ, Pomper MG, Pletnikova O, Troncoso JC, Ko HS, Dawson VL, Dawson TM, Lee S. Blocking microglial activation of reactive astrocytes is neuroprotective in models of Alzheimer's disease. Acta Neuropathol Commun. 2021 Apr 26;9(1):78. PubMed.

   Kampmann M. CRISPR-based functional genomics for neurological disease. Nat Rev Neurol. 2020 Sep;16(9):465-480. Epub 2020 Jul 8 PubMed.

   Herrmann JE, Imura T, Song B, Qi J, Ao Y, Nguyen TK, Korsak RA, Takeda K, Akira S, Sofroniew MV. STAT3 is a critical regulator of astrogliosis and scar formation after spinal cord injury. J Neurosci. 2008 Jul 9;28(28):7231-43. PubMed.

   Ben Haim L, Ceyzériat K, Carrillo-de Sauvage MA, Aubry F, Auregan G, Guillermier M, Ruiz M, Petit F, Houitte D, Faivre E, Vandesquille M, Aron-Badin R, Dhenain M, Déglon N, Hantraye P, Brouillet E, Bonvento G, Escartin C. The JAK/STAT3 pathway is a common inducer of astrocyte reactivity in Alzheimer's and Huntington's diseases. J Neurosci. 2015 Feb 11;35(6):2817-29. PubMed.

   Reichenbach N, Delekate A, Plescher M, Schmitt F, Krauss S, Blank N, Halle A, Petzold GC. Inhibition of Stat3-mediated astrogliosis ameliorates pathology in an Alzheimer's disease model. EMBO Mol Med. 2019 Feb;11(2) PubMed.

   Anderson MA, Burda JE, Ren Y, Ao Y, O'Shea TM, Kawaguchi R, Coppola G, Khakh BS, Deming TJ, Sofroniew MV. Astrocyte scar formation aids central nervous system axon regeneration. Nature. 2016 Apr 14;532(7598):195-200. Epub 2016 Mar 30 PubMed.

   Rubio N, Sanz-Rodriguez F, Arevalo MA. Up-regulation of the vascular cell adhesion molecule-1 (VCAM-1) induced by Theiler's murine encephalomyelitis virus infection of murine brain astrocytes. Cell Commun Adhes. 2010 Jun;17(3):57-68. PubMed.

   Labib D, Wang Z, Prakash P, Zimmer M, Smith MD, Frazel PW, Barbar L, Sapar ML, Calabresi PA, Peng J, Liddelow SA, Fossati V. Proteomic Alterations and Novel Markers of Neurotoxic Reactive Astrocytes in Human Induced Pluripotent Stem Cell Models. Front Mol Neurosci. 2022;15:870085. Epub 2022 May 3 PubMed.

   View all comments by Hanseok Ko

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Microglia Give Astrocytes License to Kill 20 Jan 2017

Further Reading

No Available Further Reading