Anti-tau VHH therapy against PHF6: a safe approach to slowing the phenotype of tau pathology

Abstract

Background: Tauopathies share common features, including tau aggregation, which plays a central role in neurodegeneration. However, these disorders are highly heterogeneous, particularly in the spread of pathological tau species between cells. In Alzheimer's disease, intracellular tau aggregation is followed by a propagation between cells leading to a hierarchical pathway of neurodegeneration, whereas in other tauopathies, such as progressive supranuclear palsy (PSP), pathological tau remains largely confined within neurons and exhibits more limited spread. This variability raises the question of whether tailored treatments for each tauopathy might offer more therapeutic benefit. Hence, we designed two different immunological approaches using single domain antibody fragments, also called VHHs, to target intracellular and extracellular tau. This study aims to first evaluate the safety of these immunological tools on physiological tau and then their potential to slow disease progression.

Methods: We selected the pro-aggregative tau hexapeptide PHF6 as a common target for the VHHs. These VHHs were cloned in viral vectors allowing to compare two different expression systems: 1) intracytosolic expression to prevent tau accumulation (intraVHH) and 2) secretion into the interstitial fluid, to prevent tau spreading (extraVHH). By stereotactic injection of viral vectors, these VHHs were expressed in the brain of transgenic or wild-type mice and three readouts were studied: behavior, brain imaging and tau lesions.

Results: We validated the correct addressing of intra- and extraVHHs. These two constructs were not associated with adverse effects, even in the absence of tau overexpression, in wild-type mice. Their efficacy was demonstrated in transgenic mouse tau models, either chronic long-term or in acute seeding with injections of human brain homogenates from Alzheimer's disease patients. They both can slow down several pathological effects (i.e. cognitive deficits, cerebral atrophy and neuronal hyperphosphorylation of tau).

Conclusions: This study is a proof of concept demonstrating that VHHs can be engineered to reduce both intra- and extracellular tau pathologies without major adverse effects, making them of interest for therapeutic applications.

Keywords: Alzheimer’s disease; Immunotherapy; PHF6; PSP; Single domain antibodies; Tau spreading; Tauopathies; VHH.

Fig. 1

VHH constructs- (A, B) AAV2/9 constructs encoding under the expression control of a CMV promoter. A intraZ70 or intraCtrl—VHHs fused to a mCherry tag protein or (B) extraZ70 or extraCtrl—VHHs fused to an IL2 secretion signal and a Fc portion of mouse IgG2 (C) HEK293T cells were transfected with intraCtrl, intraZ70, extraCtrl and extraZ70 plasmids and VHH expression was detected by Western blot in cell lysate (C) or media (M) using anti-bodies against VHHs (upper panel) or antibodies against Fc (lower panel). The black star indicates the presence of cell culture medium contaminants, which are also present in non-transfected cells (NT). The black arrows point to the VHHs. D AAVs encoding VHHs were injected into the hippocampus of Tg30, and the viral DNA episome from the hippocampus were detected by qPCR, 1 or 9 months after injection. E Immuno-fluorescence detection of extraVHH (anti-VHH) and the endoplasmic reticulum (anti-calreticulin) was done using hippocampal sagittal sections of mice, 9 months post-injection. ExtraVHH is visualized in green and the ER in purple. Fluorochrome overlays are analyzed in cells and illustrated on the right part of the figure

Fig. 2

VHH Safety: Brain imaging- (A) Scheme of the experimental design created with Biorender.com. (B) Mice were weighed monthly and the results are expressed as weight gain (= % weight compared to weight before viral vector injections). (C) 8 months after injection, MRI scans were used to determine the following image aspects: 1) brain volume: the statistical difference between mice injected with AAV2/9 and mice injected with GFP, intraZ70 or extraZ70 is shown (left to right). These statistical values are color-coded on the NUS template used for segmentation. The brain volumes expressed in mm³ are shown for the hippocampus (D) and for the visual cortex (H). 2) DTI signals (FA and MD) were quantified in the hippocampus (E, F respectively) and the visual cortex (I, J respectively). 3) Glucose uptake (FDG-PET) was also evaluated in the hippocampus (G) and visual cortex (K). Some examples of MBP and CNPase proteins detected by SDS-PAGE in hippocampal lysate are shown (L). Quantifications were done from all mice (M and N respectively). *: p < 0.05; **: p < 0.01; ****: p < 0.0001

Fig. 3

VHH Safety: Parenchymal and peripheral biomarkers- Brain lysates of 10-months-old WT mice were used to quantify hippocampal biomarkers by Western-blotting (A-I): TauCter (A), NeuN (B), PSD95 (C), Snap25 (D), GFAP (E), Iba1 (F), LC3b (G), pEIF2α (H), ATF4 (I). The list of antibodies used is listed in Table 1. Antibodies were validated against positive and negative controls, and some examples of mouse samples are shown (Fig. S5). Representative images of Iba1 immunostaining in the hippocampus are shown. Scale bars are indicated in the figure. Zoom areas (small black insert) are shown in the bottom right corner (large black insert) (J, left panels). Iba1 immunostaining was quantified blindly using Mercator software. Results are expressed as the percentage of Iba1-positive area relative to control (J, right panels) Plasma dosage using the Simoa technology are shown: Tau (K), NfL (L), GFAP (M) and UCH-L1 (N). *: p < 0.05; **: p < 0.01; ***: p < 0.001

Fig. 4

VHH Safety: behavioral assays- The distance traveled (A, B) and velocity (C, D) of 9-month-old WT mice were monitored in a 10 min OpenField area and in a 5 min Y-maze assay. E OpenField test. Time near the wall (P = periphery) or in the center (C) is reported relative to the total time in the arena (10 min). F Y-maze test. Time in new arm was reported as the sum of time in new and other arms in the first minute. The p value is related to the % of time spent in the new arm in comparison to 50% (dotted line). *: p < 0.05; **: p < 0.01; ***: p < 0.001; ****: p < 0.0001

Fig. 5

Behavioral tests—(A) Scheme of the experimental design created with Biorender.com. The motor abilities of 8-month-old Tg30 mice injected with intra- or extraVHHs, or WT mice injected with AAV2/9, were controlled by distance traveled (B, C) and the velocity (D, E) in a 10 min OpenField area or in a 5 min Y-maze assay. F OpenField test. Time spent near the wall (P = periphery) or in the center (C) is reported relative to the total time in the arena (10 min). G Y-maze test. Time in the new arm is reported relative to the sum of time in the new and other arms in the first minute. The p value is related to the % of time spent in the new arm in comparison to 50% (dotted line). F-G GraphPad Prism 8 software (version 8.0.0) was used to exclude outliers from the analysis. *: p < 0.05; **: p < 0.01; ****: p < 0.0001

Fig. 6

MRI-brain volume evaluation- (A) MRI scans taken 8 months after injection were used to determine the statistical difference in the brain volume between AAV2/9-injected WT mice and Tg30 mice injected with intraCtrl, intraZ70, extraCtrl, or extraZ70 (left to right). These statistical values were color-coded on the NUS template used for segmentation. The brain volume expressed in mm3 is shown for the hippocampus (B) or visual cortex (C), quantified by MRI. *: p < 0.05; **: p < 0.01; ***: p < 0.001, ns: non-significant

Fig. 7

Tau pathology in aged Tg30, a chronic model of tau pathology expressing intraZ70 or extraZ70- (A) Illustration of AT8 tau lesions in hippocampal sections. Scale bars are indicated in the figures. Immunoreactivity is detected and quantified using the AT8 antibody in the hippocampus (B). Immunohistofluorescence was done on sagittal brain sections to visualize phosphoTau (AT8, green) and VHH (purple). The specific location of phosphoTau and intraZ70 (C) or intraCtrl (D) was quantified across the cells (left profile). Fluorochrome overlays are analyzed in cells and illustrated on the right part of the figure. Quantification of MC1 (E) or AT100 (F) immunostaining in the hippocampus is also shown. For B, E and F, quantifications were done using Mercator software. Results are presented as the percentage of antibody stained surface relative to the control. **: p < 0.01

Fig. 8

Tau pathology in young Tg30, an acute in vivo model of tau seeding expressing extraZ70- (A) Scheme of the experimental design created with Biorender.com. B Representative images of mice with AT8 immunostaining (AT8 intensity closest to the group mean value). The needle track is indicated with a black arrow. Scale bars are indicated in the figure. C AT8 immunoreactivity in the whole hippocampus were blindly quantified using the Mercator software. Results are expressed as the percentage of AT8-positive area relative to control. *: p < 0.05