Species: Mouse
Genes: Trem2
Modification: Trem2: Knock-In
Disease Relevance: None
Strain Name: B6-Trem2em3Npa
Genetic Background: C57BL/6J
Availability: Available under MTA from Novartis Pharma AG. Contact Derya Shimshek (derya.shimshek@novartis.com).

Summary

Triggering Receptor Expressed on Myeloid Cells 2 (TREM2) is a transmembrane receptor found on microglia, where it modulates cell activity and survival. In addition to membrane-bound TREM2, there are soluble forms of the protein—generated by protease cleavage of the extracellular domain or expression of alternative transcripts that lack a transmembrane domain.

TREM2 is cleaved by ADAM proteases after histidine 157 to release a soluble N-terminal fragment (Feuerbach et al., 2017; Schlepckow et al., 2017; Thornton et al., 2017; see Aug 2017 news). In an attempt to create a model that generates only soluble TREM2 (sTREM2)—but no full-length, signaling-competent receptor—CRISPR/Cas9 gene editing was used to introduce a stop codon after H157 of murine Trem2 (Beckmann et al., 2023). These mice, called “TREM2-sol,” were found to express very low levels of sTREM2 and Trem2 mRNA. Nonetheless, differences between TREM2-sol and Trem2 knock-out mice (Trem2-KO) were observed, including prolonged microglial responses to injury, increased vulnerability of bone marrow-derived macrophages to growth factor deprivation, and preservation of endo-lysosomal function in TREM2-sol compared with Trem2-KO mice.

The following description refers to mice homozygous for the Trem2 mutant allele.

Levels of Trem2 expression | Effects on myeloid cells in vitro | Pathological context | Modification details | Related Models

Levels of Trem2 expression

As mentioned above, TREM2-sol mice generated very little sTREM2, compared with mice that express wild-type TREM2. Levels of TREM2 protein in the forebrains of TREM2-sol mice were only about 20 percent of those measured in wild-type mice, when extracted using conditions expected to enrich for sTREM2 over membrane-associated TREM2. Trem2 mRNA levels were also found to be much lower in in the forebrains of TREM2-sol mice, compared with wild-type mice, although it should be noted that protein levels were measured in 4- to 5-month-old animals, while mRNA was measured in 12-month mice. The reason for the reduced expression of Trem2 in TREM2-sol mice was not pursued, but nonsense-mediated decay is a possibility.

Additionally, levels of sTREM2 measured in conditioned media from bone marrow-derived macrophages (BMDM) cultured from TREM2-sol mice were about one-tenth those from wild-type BMDM. No TREM2 was detected on the surface of BMDM from TREM2-sol mice.

It is not clear whether the sTREM2 measured in these mice represents TREM2 amino acids 1-157. Soluble TREM2 in mice can be generated by cleavage of TREM2 derived from the canonical transcript (ENSMUST00000024791), which encodes a 227-amino acid isoform containing a transmembrane domain. Mouse TREM2²²⁷, like the canonical, 230-amino acid form of human TREM2, has an ADAM protease cleavage site after histidine-157. TREM2²²⁷ also has a meprin β cleavage site after arginine-136, and this cleavage has been observed on macrophages (Berner et al., 2020), although it has not yet been reported for microglia. Additionally, a soluble form of murine TREM2 can be generated by expression of an alternate transcript (ENSMUST00000113237; Moutinho et al., 2023). This transcript has a 55-base pair insertion between exons 3 and 4 and encodes a 249-amino acid isoform lacking a transmembrane domain (Schmid et al., 2002). Cultured microglia and BMDM express this alternate transcript at about one-tenth the level of the canonical transcript (Schmid et al., 2002). As in many other studies, the antibodies used by Beckmann and colleagues to detect sTREM2 protein would not distinguish between soluble cleavage products and the soluble isoform resulting from alternative splicing. Nor would the primers used to measure mRNA distinguish between the canonical and alternate transcripts.

Effects on myeloid cells in vitro

BMDM cultured from TREM2-sol mice were more vulnerable to M-CSF (macrophage colony stimulating factor) deprivation than BMDM from either wild-type or Trem2-KO mice. Two days after withdrawal of M-CSF, the survival rate of wild-type BMDM was about 80 percent, while the survival rates for Trem2-KO and TREM2-sol were 70 percent and 25 percent respectively.

Phagocytic activity, assessed as the uptake of pHrodo-myelin (myelin labeled with a pH-sensitive fluorescent tag), was impaired in microglia and BMDM from TREM2-sol mice. Uptake of pHrodo-myelin by TREM2-sol and Trem2-KO microglia was about one-third that of wild-type microglia. Phagocytic activity of BMDM was less severely impaired, with pHrodo-myelin uptake in TREM2-sol and Trem2-KO BMDM about 65 percent and 80 percent of wild-type, respectively.

Endo-lysosomal function of BMDM from TREM2-sol mice appeared normal, in contrast to BMDM from Trem2-KO mice. Endo-lysosomal activity was assessed using Magic red, a membrane-permeant dye that fluoresces after cleavage by Cathepsin B inside lysosomes. In cells treated with a compound that inhibits the fusion of lysosomes and autophagosomes, Magic red accumulated inside lysosomes. While the fluorescent signals in TREM2-sol and wild-type BMDM were comparable under these conditions, the signal from Trem2-KO cells was reduced by about half.

Pathological context

Demyelination was induced by feeding mice the copper chelator cuprizone. The results described here refer to the external capsule, as this white-matter tract was particularly affected. Five-week administration of cuprizone caused demyelination and loss of oligodendrocytes in wild-type, TREM2-sol and Trem2-KO mice, with the latter group most severely affected. While myelin began to recovery in wild-type mice during the four-week period following discontinuation of cuprizone, MRI studies suggested further demyelination during this phase in the other two genotypes, albeit at a slower rate than during cuprizone intoxication. Oligodendrocyte numbers began to recover in wild-type and TREM2-sol, but not Trem2-KO, mice when cuprizone was discontinued.

Myelin debris was not seen in wild-type mice, but it did accumulate in TREM2-sol and Trem2-KO animals during the cuprizone phase. During the recovery phase, the amount of myelin debris declined in TREM2-sol, but not Trem2-KO, mice. These observations suggest that, while wild-type microglia were able to efficiently clear degenerating myelin, TREM2-sol and Trem2-KO microglia were not, consistent with the finding of impaired pHrodo-myelin uptake by TREM2-sol and Trem2-KO microglia in vitro.

During the period of cuprizone administration, wild-type, TREM2-sol, and Trem2-KO mice all showed signs of microgliosis: Microglial numbers increased, microglial morphology changed from a ramified to a more amoeboid shape, and staining for the lysosomal marker LAMP1 increased. While these indicators began to return towards baseline during the recovery period in wild-type and Trem2-KO mice, they further increased in TREM2-sol animals. Astrogliosis—an increase in the number of GFAP-positive cells—developed similarly in the three genotypes during cuprizone administration. Astrocyte numbers in wild-type mice decreased towards baseline during the recovery phase, but they remained elevated in TREM2-sol and Trem2-KO mice.

Five weeks of cuprizone administration did not appear to cause axon damage in wild-type mice but did result in damage in TREM2-sol and Trem2-KO mice that continued through the recovery phase: There was a partial loss of SMI312-immunoreactive neurofilaments in the external capsule in TREM2-sol and Trem2-KO mice, which persisted (TREM2-sol) or worsened (Trem2-KO) during the recovery phase. Levels of neurofilament light chain (NfL) in plasma, a marker of neuron injury, also increased during cuprizone administration in TREM2-sol and Trem2-KO mice, and continued to rise even after cuprizone was discontinued.

Modification details

CRISPR/Cas9 gene editing was used to introduce a stop codon after H157 of murine Trem2.

Related Models

The following models have been used to manipulate levels of soluble TREM2, through genetic alteration of the ADAM protease cleavage site or AAV-mediated expression of an extracellular fragment of TREM2.

Trem2-H157Y knock-in. CRISPR/Cas9 gene editing was used to introduce the H157Y mutation into the mouse Trem2 gene. This variant first gained attention when it was found to associate with an increased risk for Alzheimer’s disease in a Han Chinese cohort (Jiang et al., 2016). Elevated levels of sTREM2 were found in mice homozygous for the variant. The H157Y mutation did not affect microglial density or morphology, performance on a battery of behavioral tests, or levels of synaptic markers, but enhanced hippocampal synaptic plasticity. Perhaps surprisingly, when Trem2-H157Y mice were crossed with 5xFAD mice, the Trem2 variant decreased amyloid pathology, microgliosis, and plaque-associated neuritic damage.

Trem2-H157Y x 5xFAD. Trem2-H157Y mice were intercrossed with 5xFAD mice, a model of aggressive amyloidosis. Levels of Trem2 mRNA and TREM2 protein—both full-length, membrane-associated TREM2 and sTREM2—were lower in 5xFAD mice homozygous for Trem2-H157Y, compared with 5xFAD mice expressing wild-type Trem2. This difference may reflect the lower number of microglia in the brains of the Trem2 mutation carriers. However, the ratio of sTREM2 to full-length TREM2 was higher in the Trem2 mutation carriers, consistent with increased shedding of TREM2-H157Y.

Compared with 5xFAD mice homozygous for wild-type Trem2, mice homozygous for the H157Y variant showed age-dependent decreases in plaque burdens, microgliosis and plaque-associated neuritic damage. At 8.5 months, the only timepoint reported to date, expression of neuroinflammation-related genes was downregulated in the H157Y mutation carriers.

TREM2-IPD. CRISPR/Cas9 gene editing was used to disrupt the ADAM10/17 recognition site on mouse TREM2, changing histidine-157 to isoleucine (I), serine-158 to proline (P), and threonine-159 to aspartate (D)—leading to both a reduction in sTREM2 and an increase in cell-surface TREM2. The IPD mutation accelerated microglial maturation and increased microglial phagocytic activity. In the cuprizone model of demyelination, the mutation resulted in persistent neuroinflammation during the recovery phase.

TREM2-IPDxAPP23xPS45. To study the effects of reducing TREM2 cleavage in the context of amyloid pathology, TREM2-IPD mice were intercrossed with APP23 and PS45 mice (Herzig et al., 2004), which carry transgenes for human APP with the AD-linked Swedish mutation and PSEN1 with the AD-linked G384A mutation, respectively. Compared with APP23xPS45 mice carrying wild-type Trem2, TREM2-IPDxAPP23xPS45 mice had higher plaque burdens and more severe plaque-associated pathology at an early—but not late—stage of plaque deposition.

AAV-sTREM2 5xFAD. AAV-sTREM2 5xFAD mice were created to study the long-term effects of soluble TREM2 (sTREM2) in the context of amyloidosis. To generate this model, AAV carrying cDNA encoding EGFP- and FLAG-tagged sTREM2 (amino acids 1-171 of human TREM2) was injected into the brains of neonatal 5xFAD mice. Overexpression of sTREM2 led to decreased plaque burdens, increased the number of plaque-associated microglia, and rescued hippocampal long-term potentiation and performance in the Morris water maze in 5xFAD mice aged 6 to 7 months.

AAV-sTREM2 PS19. AAV-sTREM2 PS19 mice were created to study the long-term effects of soluble TREM2 (sTREM2) in the context of tauopathy. To generate this model, AAV carrying cDNA encoding EGFP- and FLAG-tagged sTREM2 (amino acids 1-171 of human TREM2) was injected into the brains of 3-month-old PS19 mice. AAV-mediated expression of sTREM2 protected against hippocampal synapse loss, improved performance in the Morris water maze and Y-maze, enhanced long-term potentiation, and reduced levels of p-tau202 and p-tau396 in PS19 mice studied at 7 months of age.

Phenotype Characterization

Complementary Models

TREM2-sol mice are genetically modified to produce only soluble TREM2 (sTREM2) but no signaling-competent cell-surface TREM2. Theoretically, this model could be used to determine which TREM2-dependent functions are supported by soluble forms of the protein. A limitation of the TREM2-sol mouse is its very low expression of TREM2. In a second genetically modified mouse line, Trem2-H157Y knock-in, levels of sTREM2 are elevated in the presence of cell-surface TREM2.

Another approach to investigating the role of sTREM2 is the direct application of the protein to animals or cultured cells. This was the strategy employed by Guojun Bu, Xiao-Fen Chen and colleagues, who used  a recombinant, chimeric protein intended to mimic sTREM2 (Zhong et al., 2017; 24 Feb 2017 news). sTREM2-Fc consists of the TREM2 extracellular domain (amino acids 1-171) with human IgG-Fc fused to its C-terminal to aid in purification of the recombinant protein. (At the time these studies began, it was not yet known that TREM2 was cleaved after amino acid 157 to generate sTREM2.) Application of IgG-Fc alone served as a control in these studies.

When applied to microglia cultured from neonatal wild-type or Trem2 knock-out mice, sTREM2-Fc promoted cell survival, induced morphological changes characteristic of “activated” microglia (enlarged cell bodies, amoeboid shape), and increased expression of inflammatory cytokines. Human TREM2 (amino acids 1-171) without Fc fusion and mouse sTREM2-Fc also promoted microglial survival.

The acute effects of human sTREM2-Fc in vivo were examined 24 hours after injection of the chimeric protein into the hippocampi of 7- to 8-month-old wild-type or Trem2 knock-out mice. Here, too, sTREM2-Fc induced a change in microglial morphology from a ramified to an amoeboid shape and increased the expression of inflammatory cytokines.

Notably, contrary to the effects of exogenous sTREM2-Fc, Bu and colleagues did not observe changes in microglial morphology or upregulation of inflammatory cytokines in Trem2-H157Y knock-in mice, in which the H157Y mutation increases the generation of sTREM2 (Qiao et al., 2023).

Several factors might account for the different responses to elevated levels of sTREM2 in the Trem2-H157Y knock-in model and the sTREM2-Fc direct-application models. The presence of the Fc sequence and the extended TREM2 extracellular region of sTREM2-Fc were noted above. The duration of exposure to elevated levels of sTREM2 might be important, with sTREM2 expected to be chronically elevated in the knock-in mice. The concentrations of sTREM2 also might be critical. Bu and colleagues quantified the levels of endogenous sTREM2 in mouse brains using two different methods (Qiao et al., 2023). Using an ELISA, they measured the amount of TREM2—presumably sTREM2—in detergent-free lysates of the brains of wild-type mice and found approximately 0.4 fmol TREM2/mg brain tissue. They also immunoprecipitated TREM2 from the brains of 5xFAD mice and measured the amount of bona fide sTREM2 by mass spectrometry, finding around 15 fmol sTREM2/mg brain tissue. (Levels of total TREM2 had previously been shown to be elevated in the brains of 5xFAD mice, a model of aggressive amyloidosis, compared with wild-type mice [Wang et al., 2020]). They then retrospectively estimated the concentration of sTREM2-Fc in the hippocampi of the mice in their 2017 study. Assuming that the injected sTREM2-Fc spread throughout the entire hippocampus but did not spread much beyond this structure, they calculated that the concentration of sTREM2-Fc was about 3.3 pmol sTREM2-Fc/mg brain tissue—more than 8000 times the amount of endogenous sTREM2 measured in brain lysates from wild-type mice and more than 200-fold that of endogenous sTREM2 immunoprecipitated from the brains of 5xFAD. (It should be noted that these comparisons assume that there was a nearly complete recovery of endogenous sTREM2 from brain lysates.)

The amount of sTREM2-Fc added to cultured cells (20 nM) appears to be of the same order of magnitude as that found in the brain. If a milligram of brain tissue occupies a microliter and sTREM2 is evenly distributed in this volume, 0.4-to-15 fmol/mg brain tissue would equate to 0.4-to-15 nM. However, if sTREM2 is concentrated in the extracellular space, which accounts for only about 20 percent of the volume of the hippocampus (Zhang and Verkman, 2010), the extracellular concentration of sTREM2 would be around 2 nM to 75 nM.

Recombinant HA- and FLAG-tagged sTREM2 isoforms like those generated from protease cleavage (human and mouse HA-sTREM2¹⁵⁷-FLAG) or alternative splicing (human HA-sTREM2²²²-FLAG, hu HA-sTREM2²¹⁹-FLAG, and mouse HA-sTREM2²⁴⁹-FLAG) impaired long-term potentiation when applied to hippocampal slices from 4-month-old wild-type mice. All peptides were used at concentrations of 15 ng/ml, or about 400-700 pM (Moutinho et al., 2023). These findings contrast with those reported for Trem2-H157Y knock-in mice, where enhanced LTP was seen in in slices from 6-month-old mice homozygous for the H157Y Trem2 allele (Qiao et al., 2023).

COMMENTS / QUESTIONS

No Available Comments

Make a comment or submit a question

To make a comment you must login or register.

References

News Citations

1. TREM2 Cleavage Site Pinpointed: A Gateway to New Therapies? 31 Aug 2017
2. Does Soluble TREM2 Rile Up Microglia? 24 Feb 2017

Research Models Citations

1. Trem2 KO (KOMP)
2. Trem2-H157Y knock-in
3. Trem2-H157Y x 5xFAD
4. 5xFAD (C57BL6)
5. TREM2-IPD
6. Trem2-IPDxAPP23xPS45
7. APP23
8. AAV-sTREM2 5xFAD
9. AAV-sTREM2 PS19
10. Tau P301S (Line PS19)
11. Trem2-H157Y knock-in
12. 5xFAD (C57BL6)

Mutations Citations

1. TREM2 H157Y
2. APP K670_M671delinsNL (Swedish)
3. PSEN1 G384A

Paper Citations

1. Feuerbach D, Schindler P, Barske C, Joller S, Beng-Louka E, Worringer KA, Kommineni S, Kaykas A, Ho DJ, Ye C, Welzenbach K, Elain G, Klein L, Brzak I, Mir AK, Farady CJ, Aichholz R, Popp S, George N, Neumann U. ADAM17 is the main sheddase for the generation of human triggering receptor expressed in myeloid cells (hTREM2) ectodomain and cleaves TREM2 after Histidine 157. Neurosci Lett. 2017 Nov 1;660:109-114. Epub 2017 Sep 18 PubMed.
2. Schlepckow K, Kleinberger G, Fukumori A, Feederle R, Lichtenthaler SF, Steiner H, Haass C. An Alzheimer-associated TREM2 variant occurs at the ADAM cleavage site and affects shedding and phagocytic function. EMBO Mol Med. 2017 Oct;9(10):1356-1365. PubMed.
3. Thornton P, Sevalle J, Deery MJ, Fraser G, Zhou Y, Ståhl S, Franssen EH, Dodd RB, Qamar S, Gomez Perez-Nievas B, Nicol LS, Eketjäll S, Revell J, Jones C, Billinton A, St George-Hyslop PH, Chessell I, Crowther DC. TREM2 shedding by cleavage at the H157-S158 bond is accelerated for the Alzheimer's disease-associated H157Y variant. EMBO Mol Med. 2017 Oct;9(10):1366-1378. PubMed.
4. Beckmann N, Neuhaus A, Zurbruegg S, Volkmer P, Patino C, Joller S, Feuerbach D, Doelemeyer A, Schweizer T, Rudin S, Neumann U, Berth R, Frieauff W, Gasparini F, Shimshek DR. Genetic models of cleavage-reduced and soluble TREM2 reveal distinct effects on myelination and microglia function in the cuprizone model. J Neuroinflammation. 2023 Feb 8;20(1):29. PubMed.
5. Berner DK, Wessolowski L, Armbrust F, Schneppenheim J, Schlepckow K, Koudelka T, Scharfenberg F, Lucius R, Tholey A, Kleinberger G, Haass C, Arnold P, Becker-Pauly C. Meprin β cleaves TREM2 and controls its phagocytic activity on macrophages. FASEB J. 2020 May;34(5):6675-6687. Epub 2020 Apr 1 PubMed.
6. Moutinho M, Coronel I, Tsai AP, Di Prisco GV, Pennington T, Atwood BK, Puntambekar SS, Smith DC, Martinez P, Han S, Lee Y, Lasagna-Reeves CA, Lamb BT, Bissel SJ, Nho K, Landreth GE. TREM2 splice isoforms generate soluble TREM2 species that disrupt long-term potentiation. Genome Med. 2023 Feb 20;15(1):11. PubMed.
7. Schmid CD, Sautkulis LN, Danielson PE, Cooper J, Hasel KW, Hilbush BS, Sutcliffe JG, Carson MJ. Heterogeneous expression of the triggering receptor expressed on myeloid cells-2 on adult murine microglia. J Neurochem. 2002 Dec;83(6):1309-20. PubMed.
8. Jiang T, Tan L, Chen Q, Tan MS, Zhou JS, Zhu XC, Lu H, Wang HF, Zhang YD, Yu JT. A rare coding variant in TREM2 increases risk for Alzheimer's disease in Han Chinese. Neurobiol Aging. 2016 Jun;42:217.e1-3. Epub 2016 Mar 3 PubMed.
9. Herzig MC, Winkler DT, Burgermeister P, Pfeifer M, Kohler E, Schmidt SD, Danner S, Abramowski D, Stürchler-Pierrat C, Bürki K, van Duinen SG, Maat-Schieman ML, Staufenbiel M, Mathews PM, Jucker M. Abeta is targeted to the vasculature in a mouse model of hereditary cerebral hemorrhage with amyloidosis. Nat Neurosci. 2004 Sep;7(9):954-60. PubMed.
10. Zhong L, Chen XF, Wang T, Wang Z, Liao C, Wang Z, Huang R, Wang D, Li X, Wu L, Jia L, Zheng H, Painter M, Atagi Y, Liu CC, Zhang YW, Fryer JD, Xu H, Bu G. Soluble TREM2 induces inflammatory responses and enhances microglial survival. J Exp Med. 2017 Mar 6;214(3):597-607. Epub 2017 Feb 16 PubMed.
11. Qiao W, Chen Y, Zhong J, Madden BJ, Charlesworth CM, Martens YA, Liu CC, Knight J, Ikezu TC, Aishe K, Zhu Y, Meneses A, Rosenberg CL, Kuchenbecker LA, Vanmaele LK, Li F, Chen K, Shue F, Dacquel MV, Fryer J, Pandey A, Zhao N, Bu G. Trem2 H157Y increases soluble TREM2 production and reduces amyloid pathology. Mol Neurodegener. 2023 Jan 31;18(1):8. PubMed. Correction.
12. Wang Y, Lin Y, Wang L, Zhan H, Luo X, Zeng Y, Wu W, Zhang X, Wang F. TREM2 ameliorates neuroinflammatory response and cognitive impairment via PI3K/AKT/FoxO3a signaling pathway in Alzheimer's disease mice. Aging (Albany NY). 2020 Oct 16;12(20):20862-20879. PubMed.
13. Zhang H, Verkman AS. Microfiberoptic measurement of extracellular space volume in brain and tumor slices based on fluorescent dye partitioning. Biophys J. 2010 Aug 9;99(4):1284-91. PubMed.
14. Moutinho M, Coronel I, Tsai AP, Di Prisco GV, Pennington T, Atwood BK, Puntambekar SS, Smith DC, Martinez P, Han S, Lee Y, Lasagna-Reeves CA, Lamb BT, Bissel SJ, Nho K, Landreth GE. TREM2 splice isoforms generate soluble TREM2 species that disrupt long-term potentiation. Genome Med. 2023 Feb 20;15(1):11. PubMed.

Further Reading