Microglia keep the brain on an even keel, and in doing so, they themselves undergo profound changes, especially when neurons begin to degenerate. Alas, exactly what molecular mechanisms are driving such changes remains a crucial question. Studies have implicated ApoE and TREM2, both genetic risk factors for Alzheimer’s disease, but a study published December 12 in Alzheimer’s and Dementia now offers a subtle twist. According to researchers led by Michelle Ehrlich and Sam Gandy at the Icahn School of Medicine Mount Sinai, New York, TREM2’s binding partner TYROBP boosts ApoE independently of TREM2. This, the scientists claim, may be the initiating step in the emergence of disease-associated microglia, or DAM (Jun 2017 news). 

  • Microglial TYROBP upregulated near sites of injury in mice.
  • The TYROBP modulates ApoE mRNA transcription.
  • TYROBP-ApoE signaling does not require TREM2 and could be the initiating step in the DAM pathway.

Research has shown that triggering TREM2 induces ApoE activation, which then leads to a switch in microglia from their homeostatic to a DAM phenotype. Blocking this TREM2-ApoE pathway restores DAM microglia to a homeostatic state in models of AD and amyotrophic lateral sclerosis (Krasemann et al., 2017). Scientists have described a two-stage DAM transition; it starts with a TREM2-independent step where TYROBP is upregulated, and is followed by a TREM2-dependent step where both TYROBP and TREM2 are upregulated (Keren-Shaul et al., 2017). 

Also known as DAP12, TYROBP had been identified previously as a master regulator of the molecular networks that are perturbed in AD (May 2013 webinar). Because of this and its upregulation in the early TREM2-independent DAM stage, Gandy and Ehrlich hypothesized that TYROBP senses the early accumulation of Aβ deposits and drives the microglial phenotypic switch from homeostatic to DAM states.

Plaque control? APP/PSEN1 mice (red) had more amyloid plaques than APP/PSEN1 mice with more TYROBP in their microglia (blue). [Courtesy of Sam Gandy.]

To test this, first author Mickael Audrain and colleagues first used RNA in-situ hybridization and immunochemistry to confirm that TYROBP mRNA does, in fact, become more abundant when microglia are recruited to the site of injury in wild-type mice, in an APP/PSEN1 model of amyloidosis and in a model of tauopathy. They then isolated microglia from wild-type mice and inflamed them experimentally with lipopolysaccharide. TYROBP mRNA levels didn’t budge, suggesting that its transcription intensifies only if microglia are recruited to sites of injury.

What does elevated TYROBP at those sites do to the microglial phenotype? The authors created a new mouse model called Iba1Tyrobp, in which the Iba1 promoter drives up expression of TYROBP in microglia by 2.5-fold. They crossed these animals with APP/PSEN1 mice. Their offspring had only half as many amyloid plaques by four months of age as APP/PSEN1 controls. Though this difference was statistically significant, other researchers noted that the variance in plaque number was extremely large, and questioned the meaning of the data (see image above).

Next, Audrain crossed the Iba1Tyrobp model with MAPTP301S mice. Western blot analysis of extracts from the cortex of 4-month-old animals indicated more phosphorylated tau in male and female crosses than in MAPTP301S controls, and immunohistochemistry of the cortex confirmed this (see image below).

More p-tau. MAPTP301S mice that overexpressed TYROBP (right) had more p-tau in their cortex than MAPTP301S controls (left).

To look for changes in TREM2 and ApoE, the authors used an injury model—sticking a thin needle into the brains of wild-type mice—to beckon and rile up microglia. In the brain’s intact hemisphere, most ApoE mRNA was produced by astrocytes, the main source of ApoE in the brain. On the injured side, ApoE mRNA was dramatically higher in the microglia recruited to the wound. By contrast, in TYROBP knockouts neither microglia in the control side of the brain nor those near the injury expressed ApoE. However, in TREM2 knockouts, TYROBP and ApoE expression were highly upregulated in microglia drawn to the injury compared to microglia on the intact side. The authors believe that TYROBP upregulation in microglia is independent of TREM2, and that the rise in ApoE expression depends on TYROBP.

What about plaque-associated microglia? To investigate, the authors used in-situ hybridization and immunohistochemistry to measure TYROBP and IBA1 expression in the vicinity of plaques in TgCRND8 transgenic mice, which overexpress mutant human APP fivefold. Audrain crossed these mice with TREM2 knockouts. In the offspring, TYROBP and ApoE expression was higher in microglia near plaques. To test if the ApoE increase depended on TYROBP, the authors turned to APP/PSEN1 mice crossed with TYROBP knockouts. They found a substantial decrease in ApoE expression in plaque-recruited microglia in these TYROBP nulls, compared to controls.

Audrain and colleagues consider these findings compelling evidence that upregulation of TYROBP mRNA is an early event in AD pathogenesis. They suspect it is triggered by an unknown sensing receptor when microglia rush to sites of injury or to amyloid plaques (see image below). They also believe that the TYROBP-ApoE signaling pathway works independently of TREM2.

Don’t Need You, TREM2. In response to a stab injury, Aβ deposits, or p-tau, recruited microglia turn up TYROBP transcription (left). This is triggered by some unknown “sensing receptor,” and can happen via multiple signaling pathways even in the absence of TREM2 (right). The signaling downstream of TYROBP upregulation goads homeostatic microglia to become DAM. [Courtesy of Sam Gandy.]

Others, including Christian Haass from the Ludwig Maximilian University in Munich, were skeptical. “This sensing population of microglia that emerges before the DAM population makes sense,” said Haass. “But they did not try to isolate these microglia and characterize them by transcriptomic and proteomic studies, nor did they prove their hypothesis by investigating TYROBP/ApoE single- and double-KO mice. So, their model remains quite speculative and needs much more data to support it.” Gandy noted that, because DAM microglia are located in the immediate proximity of amyloid plaques, neither bulk- nor single-cell-RNA sequencing can distinguish their transcriptomes. Spatial transcriptomics might be an option (Jul 2020 news). 

Oleg Butovsky at Brigham and Women’s Hospital, Boston, disagrees with the authors’ assessment. His group did sort amyloid-plaque-associated microglia from other microglia in the same brain tissue (Krasemann et al., 2017). “I think the paper has reached unsupported conclusions based on their analysis and inadequate experiments,” Butovsky told Alzforum.

Haass stressed that TREM2 is still present in the mice with overexpressed TYROBP. “They cannot make conclusions from these experiments on a TREM2-independent mechanism,” he said. “Also, their claim that the DAM1 transition is TREM2-independent but needs TYROBP and ApoE must be proven by TYROBP and ApoE knockouts followed by single-cell sequencing,” he wrote.

The authors know their paper is controversial and recognize that the DAM pathway is too complicated to be governed solely by TREM2 or TYROBP. However, “TYROBP is the adaptor of TREM2—whatever TREM2 does is through TYROBP, and this is important to keep in mind,” said Audrain. “We urge researchers to not focus only on TREM2. By doing so, we are probably missing some valuable insights.”—Helen Santoro

Comments

  1. We appreciate Helen Santoro’s thorough summary. We are writing to clarify comments attributed to Christian Haass that may not have considered the totality of key data upon which we base our conclusions:

    (1) In paragraph 10, Haass is quoted as saying “…nor did they prove their hypothesis by investigating TYROBP/ApoE single- and double-KO mice…” On the contrary, the molecular characterization of microglial phenotypes from TYROBP KO mice forms the underpinning for our conclusion that there exists a TYROBP-dependent pathway that can mediate the transformation of homeostatic microglia to disease-associated microglia (DAM) in the absence of TREM2.

    (2) In paragraph 12, “… Haass stressed that TREM2 is still present in the mice with overexpressed TYROBP…” Our evidence for the existence of a TREM2-independent pathway in microglia is derived from studies of TREM2 knockout mice, not from data using TYROBP-overexpressing (Iba1Tyrobp) mice. Endogenous TREM2 is present in TYROBP-overexpressing (Iba1Tyrobp) mice, and their phenotype is what led us to further investigate the TYROBP and TREM2 KO mice.

    We are in complete agreement that “the DAM pathway is too complicated to be governed solely by TREM2 or TYROBP.”

    The intent of our paper is to highlight the possible existence of roles for microglial TYROBP that are demonstrable even in the brains of TREM2 knockout mice.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Hot DAM: Specific Microglia Engulf Plaques
  2. Paper Alert: Those PIGs! Spatial Transcriptomics Add Human Data

Webinar Citations

  1. Can Network Analysis Identify Pathological Pathways in Alzheimer’s

Research Models Citations

  1. APPPS1
  2. TgCRND8

Mutations Citations

  1. MAPT P301S

Paper Citations

  1. . The TREM2-APOE Pathway Drives the Transcriptional Phenotype of Dysfunctional Microglia in Neurodegenerative Diseases. Immunity. 2017 Sep 19;47(3):566-581.e9. PubMed.
  2. . 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.

Further Reading

Papers

  1. . Knockdown of astrocytic TREM2 in the hippocampus relieves cognitive decline in elderly male mice. Behav Brain Res. 2021 Jan 15;397:112939. Epub 2020 Sep 28 PubMed.
  2. . Correction: Integrative approach to sporadic Alzheimer's disease: deficiency of TYROBP in cerebral Aβ amyloidosis mouse normalizes clinical phenotype and complement subnetwork molecular pathology without reducing Aβ burden. Mol Psychiatry. 2019 Mar;24(3):472. PubMed.

Primary Papers

  1. . Reactive or transgenic increase in microglial TYROBP reveals a TREM2-independent TYROBP-APOE link in wild-type and Alzheimer's-related mice. Alzheimers Dement. 2020 Dec 12; PubMed.