How TREM2 Antibodies Activate Microglia in Brain Disease

TREM2 antibodies are laboratory-engineered proteins designed to latch onto a receptor called TREM2 that sits on the surface of the brain’s resident immune cells, microglia. By switching that receptor on, or by preventing the cell from shedding it, these antibodies aim to revive microglia that have gone sluggish in diseases like Alzheimer’s. The concept has moved from mouse studies into human clinical trials over the past few years, with mixed results so far and a growing list of applications that now stretches beyond neurodegeneration into cancer immunotherapy.

Why TREM2 Matters for Brain Disease

Microglia are the brain’s cleanup crew. They engulf dead neurons, clear protein debris, and help maintain the chemical environment neurons need to function. TREM2 is a receptor on their surface that acts like a switch: when something binds it, the microglia become more active, more mobile, and better at swallowing harmful material. In Alzheimer’s disease, rare mutations in the gene that codes for TREM2 roughly triple a person’s risk of developing the condition, which is one of the strongest genetic risk factors known outside of APOE4. That discovery, made in 2013, kicked off an intense effort to find drugs that could turn TREM2 signaling back up.

The problem in Alzheimer’s is not just that toxic amyloid plaques and tangled tau protein accumulate. It is also that microglia seem to lose their ability to respond effectively. They cluster around plaques but fail to clear them. Boosting TREM2 signaling is one strategy to restore that response, and antibodies are a natural tool for the job because they can be engineered to bind a specific target with high precision.

How TREM2 Antibodies Activate Microglia

When a TREM2 antibody locks onto the receptor on a microglial cell, it triggers a cascade of internal signals. One well-characterized monoclonal antibody was shown to promote microglial proliferation, uptake of toxic amyloid-beta oligomers and dying neurons, and downstream activation through the Syk and Akt signaling pathways.

Think of it as an antibody pressing the accelerator pedal on a stalled engine. The microglia ramp up their housekeeping duties: they migrate toward amyloid plaques, wrap around them more tightly, and physically consume the debris.

Not all TREM2 antibodies work the same way. Some activate the receptor directly by clustering it on the cell surface. Others take an indirect approach by preventing the receptor from being clipped off the cell membrane, a process called shedding. When TREM2 is shed, its extracellular portion floats away into the fluid surrounding the brain, where it can be measured as “soluble TREM2” (sTREM2). That shed fragment is no longer useful for signaling. An antibody called 4D9 targets the stalk region of TREM2 near the cleavage site, stabilizing the receptor on the surface and simultaneously activating downstream signaling through phospho-SYK.

Where Antibodies Bind and Why It Matters

The exact spot where an antibody grabs TREM2 turns out to be surprisingly important. Structural studies using X-ray crystallography have shown that different antibody fragments bind to regions of TREM2 far from the natural ligand-binding site. Two fragments, scFv-2 and scFv-4, were found to attach to the opposite end of the receptor, where TREM2’s protein backbone splays open between specific structural strands to create an extended binding surface. Despite binding in a similar area, the two fragments recognized only partially overlapping patches on TREM2.

One crucial finding from structural work is that antibody fragments need to form dimers or higher-order clusters to effectively block shedding. Two fragments, scFv-3 and scFv-4, reduced sTREM2 production when applied to cells, but when researchers separated the dimers from the monomers, the single molecules had minimal or no activity. The dimers cut sTREM2 production by more than half, with maximum effects plateauing between a 50% and 70% reduction.

⁴

This cross-linking requirement has practical implications for drug design: a simple two-armed antibody may not cluster enough TREM2 molecules together to produce a strong effect, pushing engineers toward formats with more binding arms.

Engineering More Potent Formats

Standard antibodies are bivalent, meaning they have two identical binding arms. For TREM2 activation, that sometimes is not enough. Researchers at Genentech engineered a tetravalent format, doubling the binding arms on a single molecule, and found that the resulting antibody was roughly 100-fold more potent at triggering microglia to engulf amyloid-coated particles compared with the standard two-armed version. In one assay measuring phagocytosis of amyloid-beta oligomers mixed with lipids, the tetravalent molecule reached its half-maximal effect at about 0.3 nanomolar, compared with over 200 nanomolar for the conventional format.

⁵

When this tetravalent antibody was coupled with a module that hitches a ride on the transferrin receptor to cross the blood-brain barrier and then tested in a mouse model of amyloid pathology, it significantly boosted microglia-plaque interactions and enhanced the cells’ ability to swallow plaques.

⁶

Getting Past the Blood-Brain Barrier

Antibodies are large molecules, and the brain has a tightly sealed lining of blood vessels that blocks most of them from entering. Only a tiny fraction of a standard antibody injected into the bloodstream actually reaches the brain, which is a serious obstacle for any antibody therapy targeting a brain receptor. One widely studied workaround is to engineer a second binding arm onto the antibody that recognizes the transferrin receptor, a protein that naturally shuttles iron across the barrier. By piggybacking on this transport system, the antibody is ferried into the brain at much higher concentrations.

PET imaging experiments confirmed the scale of improvement. When a TREM2 antibody was fused to a transferrin-receptor-binding module and labeled with a radioactive copper tracer, brain uptake in mice expressing the human transferrin receptor was up to 4.6-fold higher than the same antibody without the transport module.

⁷

A separate review of this bispecific approach reported roughly a 10-fold increase in brain delivery, with the transferred antibody reducing amyloid burden and improving cognitive function and synaptic markers in mice.

⁸

These delivery strategies are not yet in human trials for TREM2 antibodies specifically, but the same transferrin receptor shuttle concept is being tested in clinical programs for other brain targets. It is likely that next-generation TREM2 antibodies entering the clinic will incorporate some form of brain-penetrant engineering.

Effects Beyond Amyloid: Tau Pathology

Alzheimer’s disease involves two major types of protein pathology: amyloid plaques outside cells and tangled tau inside cells. Most early TREM2 antibody work focused on amyloid clearance, but newer studies have addressed the tau side of the equation. In a tauopathy mouse model, an agonistic TREM2 antibody called Ab-T1 enhanced microglial uptake of misfolded tau, reduced tau phosphorylation and seeding in neurons derived from human-tau transgenic mice, and dampened the spread of tau pathology after tau was injected directly into the brain.

⁹

This matters because some researchers worry that clearing amyloid plaques alone may not be enough to slow cognitive decline once tau pathology is well established. An antibody that can simultaneously support amyloid clearance and limit tau spread would be a much more attractive candidate. Whether that dual benefit holds up in humans remains an open question, but the preclinical data are encouraging.

Rescuing Function Lost to Genetic Variants

The R47H variant of TREM2 is one of the strongest known genetic risk factors for late-onset Alzheimer’s. People carrying this mutation have a TREM2 receptor that still reaches the cell surface but signals weakly. In lab studies, microglia and macrophages from mice engineered with the R47H variant showed survival defects and reduced migration. Treating these cells with a TREM2-activating antibody rescued the survival deficit in both young and aged animals, and the antibody also boosted migration toward a chemical attractant, with or without amyloid-beta present.

¹⁰

Interestingly, biophysical studies have shown that the R47H mutation only modestly reduces TREM2’s ability to bind one of its natural ligands, ApoE, while mutations at a different surface on the protein, a hydrophobic patch, reduced ApoE binding by 5- to 200-fold. This suggests the R47H problem is more about signaling efficiency than ligand binding per se, which is good news for the antibody approach: if the receptor can still grab its targets but just signals poorly, an antibody that amplifies signaling could compensate for the genetic defect without needing to restore the binding surface itself.

AL002 in Human Trials

The furthest-along TREM2 antibody in human testing is AL002 (latozinemab), developed by Alector and partnered with AbbVie. In a phase 1 trial of 64 healthy volunteers, a single intravenous dose of AL002 produced dose-dependent reductions in soluble TREM2 in cerebrospinal fluid, along with increases in biomarkers of TREM2 signaling and microglial recruitment. No treatment-related serious adverse events occurred over 12 weeks of follow-up.

¹¹

The phase 2 trial, called INVOKE-2, enrolled 381 participants with early Alzheimer’s disease who received AL002 at one of three dose levels or placebo intravenously every four weeks for 48 to 96 weeks. The antibody sustained target engagement in the central nervous system, as measured by reductions in soluble TREM2 and increases in osteopontin in cerebrospinal fluid.

¹²

However, INVOKE-2 did not demonstrate significant clinical benefits on cognitive outcomes, a disappointing result that has prompted the field to re-examine key variables including treatment timing, dose optimization, patient genetic backgrounds, and whether TREM2 agonism needs to be paired with other therapies.

¹³

It is worth noting that “hitting the target” and “helping the patient” are two different things. AL002 clearly reached TREM2 in the brain and triggered the expected biomarker changes. The failure was downstream: the biological response either was not large enough, not early enough, or not the right kind of activation to translate into measurable cognitive improvement over the trial period. That distinction matters because it leaves the door open for next-generation molecules, different dosing regimens, or different patient populations.

Iluzanebart and a Rare Neurological Disease

TREM2 antibodies are not only being tested in Alzheimer’s. Vigil Neuroscience developed iluzanebart (originally called VGL101), a fully human TREM2 agonist antibody aimed at adult-onset leukoencephalopathy with axonal spheroids and pigmented glia, a rare and progressive brain disease caused by mutations in the CSF1R gene.

¹⁴

In this disease, microglia are directly impaired by the genetic defect, and boosting TREM2 signaling is hypothesized to compensate for the lost CSF1R function.

Phase 1 results in healthy volunteers showed durable, dose-dependent target engagement: soluble TREM2 dropped in cerebrospinal fluid while soluble CSF1R and osteopontin levels rose, indicating changes in microglial activity after both single and repeated dosing.

¹⁵

The rare-disease context is strategically interesting because the patient population is smaller and more genetically defined, potentially making it easier to demonstrate a benefit than in a heterogeneous condition like Alzheimer’s.

TREM2 Antibodies in Cancer Immunotherapy

While neurodegeneration researchers want to activate TREM2 on microglia, cancer immunologists have found a reason to do roughly the opposite in tumors. Tumors are infiltrated by macrophages that often express TREM2 and take on an immunosuppressive character, shielding the cancer from the body’s immune response. In this setting, anti-TREM2 antibodies engineered to deplete those suppressive macrophages have shown striking results in mice.

An Fc-engineered anti-TREM2 antibody administered to tumor-bearing mice blunted tumor growth on its own and, when combined with a suboptimal dose of anti-PD-1 checkpoint therapy, achieved complete tumor control in all mice tested.

¹⁶

A separate study confirmed that an afucosylated anti-TREM2 antibody, designed to enhance its ability to kill target cells through antibody-dependent cellular cytotoxicity, progressively depleted TREM2-expressing tumor-associated macrophages and shifted the remaining macrophages toward a pro-inflammatory state. When combined with anti-PD-1 treatment, selective depletion of immunosuppressive macrophages and phenotypic shifts in the survivors made the tumor microenvironment significantly more hostile to cancer.

¹⁷

More recently, anti-TREM2 antibodies have been tested alongside neoantigen vaccines. In one study, suppressive macrophages blunted the effectiveness of a peptide-based cancer vaccine, but adding a TREM2-blocking antibody reversed that suppression, reducing the immunosuppressive macrophage population, expanding pro-inflammatory macrophages, and increasing the number of tumor-specific killer T cells.

¹⁸

The cancer application flips the TREM2 antibody concept on its head compared with Alzheimer’s. In the brain, the goal is to activate TREM2 to make microglia more protective. In tumors, the goal is to either block TREM2 or deplete the cells expressing it, because TREM2-positive macrophages in that context are helping the tumor survive. The shared biology but opposite therapeutic goals make TREM2 one of the more intellectually interesting antibody targets in medicine right now.

Combination Therapy and Unanswered Questions

If TREM2 agonism fires up microglia and anti-amyloid antibodies tag plaques for destruction, combining the two sounds logical. At least one systematic preclinical effort tested this hypothesis, combining a TREM2 agonistic antibody with an anti-amyloid-beta antibody in several treatment paradigms in mouse models. The results were deflating: no beneficial effect on amyloid levels was seen from TREM2 agonism alone or in combination with anti-amyloid treatment across one co-dosing and two priming study designs.

¹⁹

Why the combination did not work as expected is still debated. One possibility is that anti-amyloid antibodies recruit microglia through Fc-receptor signaling so effectively that adding TREM2 activation on top provides no additional push. Another is that timing matters: if the TREM2 antibody primes microglia before plaques are opsonized by the anti-amyloid antibody, the sequence may need to be reversed or the dosing window tightened. These negative preclinical results do not close the door on combination strategies, but they do complicate the “just add both” logic that initially seemed appealing.

Tracking TREM2 With PET Imaging

Beyond therapy, TREM2 antibodies have been repurposed as diagnostic tools. By attaching a radioactive tracer to a TREM2-targeting antibody and injecting it into a living animal, researchers can use positron emission tomography to map where TREM2-expressing microglia are concentrated. A recent study established quantification methods for a copper-64-labeled TREM2 antibody-based PET tracer, demonstrating noninvasive measurement of TREM2 expression in both the brain and peripheral organs of mice.

²⁰

This kind of imaging could eventually let clinicians see whether a patient’s microglia are in the right state to benefit from a TREM2 agonist before starting treatment, or to confirm that a drug is actually engaging its target in a living person’s brain rather than relying only on cerebrospinal fluid biomarkers. The fact that the same tracer detected TREM2 signals in cardiac tissue after myocardial infarction also underscores how broadly TREM2 biology extends beyond the brain, a reminder that any systemic TREM2-targeting drug will inevitably affect macrophages elsewhere in the body as well.

Microglial States and the iPSC Window

One less obvious use of TREM2 antibodies is as research tools for understanding what goes wrong in patient microglia. Using microglia derived from human induced pluripotent stem cells, researchers defined a transcriptional signature of TREM2 activation by treating the cells with an agonistic antibody and measuring which genes turned on or off.

²¹

Applying this signature to patient-derived cells revealed that microglia from Alzheimer’s patients showed reduced TREM2 activation compared with healthy controls, providing a molecular readout of the deficit the antibody therapy is trying to correct. This kind of work is important because it bridges the gap between mouse models, where TREM2 antibodies look promising, and human biology, where the results so far have been more modest. If patient microglia are fundamentally less responsive to TREM2 stimulation than mouse microglia, that could help explain why animal results have not yet translated cleanly to the clinic and guide the design of more potent next-generation molecules.