Here we describe a cell-based assay to quantitatively assess tau uptake by microglia with the aim of creating an investigational tool to better characterize the mechanisms of action of anti-tau antibodies.
Alzheimer's disease (AD) is a progressive neurodegenerative condition in which aggregated tau and amyloid proteins accumulate in the brain causing neuronal dysfunction which eventually leads to cognitive decline. Hyperphosphorylated tau aggregates in the neuron are believed to cause most of the pathology associated with AD. These aggregates are assumed to be released into the extracellular compartment and taken up by adjacent healthy neurons where they induce further tau aggregation. This "prion-like" spreading can be interrupted by antibodies capable of binding and "neutralizing" extracellular tau aggregates as shown in preclinical mouse models of AD. One of the proposed mechanisms by which therapeutic antibodies reduce pathology is antibody-mediated uptake and clearance of pathological aggregated forms of tau by microglia. Here, we describe a quantitative cell-based assay to assess tau uptake by microglia. This assay uses the mouse microglial cell line BV-2, allows for high specificity, low variability and medium throughput. Data generated with this assay can contribute to a better characterization of anti-tau antibody effector functions.
Alzheimer's disease (AD) is a progressive neurodegenerative condition characterized by the conformational change and self-assembly of amyloid β peptide and tau protein into pathological aggregates. The normal soluble amyloid β peptide is converted into oligomeric and fibrillar amyloid β, while abnormally phosphorylated tau accumulates as oligomers and neurofibrillary tangles1,2. These protein aggregates cause neuronal death leading to memory loss and subsequent progressive cognitive decline. Other factors, including non-productive neuroinflammation and a reduced ability to clear misfolded proteins, may exacerbate and accelerate disease. Currently, intervention strategies against AD provide largely symptomatic relief, but there is no disease-modifying cure or prevention.
Increasing evidence suggests a key role of hyperphosphorylated tau aggregates in the pathology of AD. In its non-pathological state, tau is a natively unfolded protein that binds to microtubules and promotes their assembly into the neuronal cytoskeleton. When tau becomes hyperphosphorylated, it detaches from the cytoskeleton and clusters into tau aggregates in the neuron, which are believed to cause most of the pathology associated with AD3. Aggregated tau starts accumulating first intracellularly, but as disease progresses, it is assumed to be released from affected neurons into the extracellular space, from which it can be taken up by adjacent or synaptically connected healthy neurons in a "prion-like manner". Once internalized, the tau aggregate induces further tau aggregation via templated conformational change4.
According to this hypothesis, therapies capable of interrupting tau seeding might slow down or reverse the course of tau-mediated neurodegenerative disease. In support of this, mice made susceptible to tauopathy by genetic mutation and passively injected with anti-tau antibodies show reduced tau pathology and improved cognitive function5,6,7,8,9. However, the mechanisms by which therapeutic antibodies reduce pathology still remain elusive.
One of the proposed mechanisms is antibody-mediated uptake and clearance of pathological aggregated forms of tau by microglia, the brain's resident immune cells. Recent publications suggest that microglia can efficiently internalize and degrade pathological tau species and this ability is enhanced by anti-tau antibodies via an Fc-dependent mechanism involving Fc receptors expressed on the surface of microglia and receptor mediated phagocytosis10,11. These data identify microglia as potentially important effectors of therapeutic antibodies.
We describe herein a cell-based assay to quantitatively assess tau uptake by microglia. Data generated with this assay can help elucidating the mechanisms of action of anti-tau antibodies thus representing a useful tool to advance anti-tau antibodies to further steps of their development as potential AD treatment.
1. BV-2 Cells Culture
NOTE: Handle BV-2 cells under Biosafety Level 2 containment. The BV-2 cell line produces an enveloped recombinant ecotropic retrovirus (capable of infecting murine cells only)12; such viruses are known for their in vitro transforming ability and in vivo tumorigenic potential.
2. Label Recombinant Tau Aggregates with pH-sensitive Fluorescent Dye
NOTE: Tau aggregates were prepared as described in Apetri et al.13 with the difference that no Thioflavin T (ThT) was added to the reaction buffer. Aggregated samples were collected in 1.5 mL centrifuge tubes. Final fluorescence signal was checked by mixing 118 µL of the pool sample with 12 µL of a 50 µM ThT solution. Aggregates were separated by centrifuging the aggregation reaction mixture at 20,000 x g for 1 h at 4 °C. The supernatant was analyzed by SEC-MALS to confirm that all the monomeric tau was converted into aggregates. Pellets (tau aggregates) were snap frozen and stored in a freezer at -80 °C.
3. Uptake Assay with Fluorescence-activated Cell Sorting (FACS) Read-out
4. FACS Analysis
NOTE: Refer to Figure 1 for the gating strategy.
5. Immunocomplexes Uptake with Microscopy Read-out
Aggregated recombinant tau was covalently labelled with a pH-sensitive green dye. This dye dramatically increases its fluorescence upon its internalization in acidic organelles, thereby allowing for intracellular quantification. Labeled tau aggregates were incubated with anti-tau monoclonal antibodies. In particular, we used a chimeric version (mouse IgG1 Fc region) of CBTAU-28.1. This human antibody binds to the N-terminal insert region of tau and is able to bind in vitro generated tau fibrils13. In this assay, we also tested an affinity-improved version of CBTAU-28.1 – dmCBTAU-28.1. Fab fragments of CBTAU-28.1, in the parental and high-affinity mutant format, and a mouse IgG1 isotype control were used as controls.
BV-2 cells were incubated with the pre-formed immunocomplexes or aggregated tau alone for two hours in the presence of heparin to block antibody-independent tau uptake. After incubation, cells were trypsinized to remove the tau bound to the extracellular membrane and were analyzed for tau uptake by flow cytometry. As we recently described13, we observed that CBTAU-28.1 variants promoted uptake of tau in BV-2 cells in a dose-dependent manner. The uptake was Fc mediated since CBTAU-28.1 Fab fragments did not increase basal tau uptake (Figure 2). Moreover, the high affinity dmCBTAU-28.1 antibody mediated tau uptake into BV-2 cells to a higher extent than the wild-type antibody (Figure 2).
Antibody-mediated tau uptake and localization of tau aggregates in the endolysosomal compartment was confirmed by confocal microscopy (Figure 3) where the acidic cellular compartment was stained using a probe selective for low pH organelles. Intracellular puncta of green pH dye labeled tau aggregates were observed inside the cells that were incubated with CBTAU-28.1. Moreover, intracellular tau aggregates often colocalized with the low pH compartment selective red dye thus suggesting presence of tau aggregates in the acidic organelles. CBTAU-28.1 Fab fragments did not increase tau uptake again indicating an Fc-receptor mediated internalization mechanism (Figure 3).
Figure 1: Gating strategy used in flow cytometry analysis to detect tau internalization by BV-2 cells. Sample data from BV-2 only control (A-C), isotype control (D-F) and dmCBTAU-28.1 (G-I) are shown. BV-2 cell population was gated on a FSC-A vs SSC-A density plot excluding debris and dead cells (A, D, G). BV-2 cells were then further gated on a FSC-A vs FSC-H density plot to exclude cell doublets and aggregates (B, E, H). Single cell gate was used to generate a pH dye (FITC in these representative results) single parameter histogram (C, F, I) and determine geometric mean fluorescence intensity. Alternatively, percentage of pH dye-tau positive cells was calculated excluding negative cells as determined by using BV-2 only control. Please click here to view a larger version of this figure.
Figure 2: CBTAU-28.1 mediates uptake of tau aggregates into microglial BV-2 cells. Aggregated recombinant tau was covalently labelled with green fluorescence pH-sensitive dye and incubated with a mouse chimeric version of the human anti-tau antibody CBTAU-28.1, its affinity improved format, dmCBTAU-28.1, the corresponding Fab fragments, a mouse IgG1 isotype control antibody or no antibody (tau aggregates alone). Immunocomplexes were subsequently incubated with BV-2 cells for two hours in the presence of heparin to block antibody-independent tau uptake. Uptake of immunocomplexes was assessed by flow cytometry and expressed as the geometric mean (GM) of fluorescence intensity (A) or percentage of tau positive (tau+) cells (B). Error bars in (A) indicate the standard deviation of two independent experiments, while (B) shows a single experiment. Please click here to view a larger version of this figure.
Figure 3: Tau aggregates are internalized by BV-2 cells and localize in cellular acidic organelles. Preformed tau-antibody immunocomplexes were incubated with BV-2 cells for two hours in the presence of heparin to block antibody-independent uptake. After incubation, nuclei were stained with a DNA specific blue dye and the acidic cellular compartment with a low pH compartment selective red dye. Live-cell imaging revealed intracellular puncta of labeled tau aggregates (green) inside the cells that were incubated with CBTAU-28.1 and dmCBTAU-28.1, but not with the isotype control. Moreover, intracellular tau aggregates often colocalized with the red dye (yellow) thus suggesting presence of tau aggregates in the acidic cellular compartment. CBTAU-28.1 Fab fragments did not increase tau uptake indicating an Fc-receptor mediated internalization mechanism. Images represent maximum intensity projections of a 20 planes Z-stack (0.5 µm planes) acquired with a 63X water immersion objective. Please click here to view a larger version of this figure.
Microglia, the resident brain's immune cells, have been recently identified as important players in antibody-mediated therapeutic approaches for tauopathies10,11. Antibody-mediated tau clearance by microglia, together with blocking of neuronal uptake9, inhibition or destabilization of fibril formation13,14 and clearance of intraneuronal fibrils via the lysosomal pathway15, might all contribute to the anti-tau antibody efficacy observed in mouse model of tauopathy5,6,7,8,9.
We described here a cell-based assay to quantitatively assess tau uptake by microglia with the aim of creating an investigational tool to better characterize the mechanisms of action of anti-tau antibodies.
This assay, adapted from Funk et al.11, uses BV-2 cells, which are immortalized murine microglial cells. While they cannot fully be compared to primary microglial cells, they feature many of the characteristics of primary microglia, including the ability of robustly phagocytose both Aβ and tau fibrils11,16,17,18,19. Moreover, they showed a reproducible behavior in vitro which made them highly suitable for assay development and quantitative studies, which require minimal experimental variability. Beside this, immortalized cell lines allow higher assay throughput and eliminate the need for animal sacrifice compared to the use of primary microglia.
The tau aggregates we used in this assay were obtained using the highly reproducible in vitro aggregation procedure that we recently described13, and show similar morphology to paired helical filaments (PHFs) isolated from brains of AD patients. While we did not observe any unexpected results that might have been caused by tau aggregates adherence to plastic or glass surfaces, the use of stable and well characterized tau aggregates played a crucial role in the reproducibility of this assay.
Another aspect that significantly contributed to assay reproducibility was cell density. The numbers of cells per well described in the protocol represent the optimal cell density in the described conditions.
Differently than what Funk et al.11 described, we labeled tau aggregates with a pH sensitive dye which significantly increases its fluorescence upon internalization in acidic organelles, thus allowing for intracellular quantification. This, together with trypsin digestion of surface bound immunocomplexes and/or tau, guarantees that fluorescence signal measured by flow cytometry is the result of tau uptake rather than binding to the cellular surface. Moreover, the use of a pH sensitive dye eases detection of internalized tau aggregates in microscopy experiments without the need of digesting surface bound immunocomplexes/tau aggregates which would then requires cell re-plating and recovery.
We also further optimized the microscopy read-out of our assay, compared to what has previously been described11, by using a highly selective dye for acidic organelles in our microscopy experiments which allowed us not only to confirm antibody-mediated tau uptake, but also localization of tau aggregates in the endolysosomal compartment.
The assay we developed, has optimal specificity which results in a good experimental window allowing a strong separation between positive and negative samples. Interestingly, the assay indirectly detects differences in antibody affinity thus representing a powerful tool to study anti-tau antibody effector functions.
The authors have nothing to disclose.
We would like to thank Alberto Carpinteiro Soares for his valuable technical assistance.
BV-2 cells | ICLC Interlab Cell Line Collection | ATL03001 | |
Phosphate Buffered Saline (PBS) (1X) | Gibco | 10010-015 | |
Trypsin-EDTA 0.05% | Gibco | 25300-054 | |
DMEM 4.5 g/dl glucose | Gibco | 41966-029 | |
Fetal Bovine Serum | Gibco | 10091-148 | |
Penicillin-Streptomycin (10,000 U/mL) | Gibco | 15140122 | |
L-Glutamine 200mM | Lonza | 17-605E | |
EasYFlask | Nunc | 156499 / 159910 | |
pHrodo Green STP ester | Life Technologies | P35369 | |
Sodium Bicarbonate pH 8.5 100 mM | |||
DMSO | Sigma | D2650-100ml | |
PD10 columns | GE Healthcare | 17-0851-01 | |
BCA Protein Assay Kit | Thermo Fisher Scientific | 23225 | |
Greiner CELLSTAR multiwell culture plates | Greiner | 665180 | |
Falcon 96-Well Assay Plates | Falcon | 353910 | |
Heparin | Sigma | H3393-50KU | |
Trypsin-EDTA 0.25% | Sigma | T4049-100ml | |
BSA | Sigma | A7030-100G | |
EDTA 0.5M, pH8 | |||
FACS Canto II | BD | ||
Hoechst 33342 Solution (20 mM) | Thermo Fisher Scientific | 62249 | |
LysoTracker Deep Red | Thermo Fisher Scientific | L12492 | |
Opera Phenix | Perkin Helmer | HH14000000 |