A monoclonal antibody TrkB receptor agonist as a potential therapeutic for Huntington's disease

Abstract

Huntington's disease (HD) is a devastating, genetic neurodegenerative disease caused by a tri-nucleotide expansion in exon 1 of the huntingtin gene. HD is clinically characterized by chorea, emotional and psychiatric disturbances and cognitive deficits with later symptoms including rigidity and dementia. Pathologically, the cortico-striatal pathway is severely dysfunctional as reflected by striatal and cortical atrophy in late-stage disease. Brain-derived neurotrophic factor (BDNF) is a neuroprotective, secreted protein that binds with high affinity to the extracellular domain of the tropomyosin-receptor kinase B (TrkB) receptor promoting neuronal cell survival by activating the receptor and down-stream signaling proteins. Reduced cortical BDNF production and transport to the striatum have been implicated in HD pathogenesis; the ability to enhance TrkB signaling using a BDNF mimetic might be beneficial in disease progression, so we explored this as a therapeutic strategy for HD. Using recombinant and native assay formats, we report here the evaluation of TrkB antibodies and a panel of reported small molecule TrkB agonists, and identify the best candidate, from those tested, for in vivo proof of concept studies in transgenic HD models.

Figure 1. TrkB signaling and assay cascades.

(A) TrkB signalling cascade showing proximal and distal assay measurement points. Red arrows represent BDNF-induced trans-phosphorylation events within the intracellular tyrosine kinase domains. Changes in TrkB phosphorylation/activation detected by (1) Invitrogen CellSensor®; (2) DiscoveRx PathHunter®; and (3) MSD® pAKT assays. (B) Screening cascade used to characterize TrkB modulators.

Figure 2. Activation of TrkB signaling pathway by TrkB mAbs.

(A) CellSensor® TrkB-NFAT-bla CHO-K1 cells were stimulated with BDNF, mAb 38B8 and mAb 29D7 ± pan-Trk inhibitor (100 nM) over the indicated concentration range for 5 hours before β-lactamase assay was performed as described in Methods. % of control (maximal BDNF concentration  =  9 nM) values were plotted for the indicated concentrations of each ligand (n = 2 ± SEM for each data point). The TrkB mAbs activated TrkB-dependent NFAT signaling that was inhibited by the pan-Trk kinase inhibitor; (B) CellSensor® TrkA-NFAT-bla, (C) CellSensor^® TrkC-NFAT-bla, and (D) CellSensor® Trk null-NFAT-bla CHO-K1 cells were stimulated with mAb 38B8, mAb 29D7 and their respective cognate ligands (TrkA:NGFβ; TrkC:NT-3; NFAT:thapsigargin) over the indicated concentration range for 5 hours before β -lactamase assay was performed as described in Methods. % of control (maximal [NGF/NT-3/thapsigargin]) values were plotted for the indicated concentrations of each ligand (n = 2 ± SEM for each data point). Data suggest that the TrkB agonist mAbs are selective and specific; (E) CellSensor® TrkB-NFAT-bla CHO-K1 cells were stimulated with 10 nM BDNF, mAb 38B8 or mAb 29D7 over a short time course (2, 5 and 10 minutes). Cell lysates were resolved by SDS-PAGE, and phosphorylation levels of TrkB, AKT and ERK1/2 were detected using phospho-specific antibodies (pTrkB Tyr516, CST#4619; pTrkB Tyr706/707, CST#4621; pTrkB, Tyr816, CST#4168; pERK1/2, CST#4051; pAKT, CST#4051). Total levels of TrkB, AKT and ERK1/2 were also assessed (TrkB, CST#4603; AKT, CST#4685; ERK1/2, CST#4695). Relevant secondary antibodies were applied and signal was detected with ECL reagent (TrkB) or using the Odyssey infra-red reader (AKT, ERK). Data suggest activation of TrkB phosphorylation signal transduction cascade by TrkB mAbs; and (F) PathHunter® U2OS TrkB cells were stimulated with BDNF, mAb 38B8 and mAb 29D7 over the indicated concentration range for 5 hours before β-galactosidase assay was performed as described in Methods. % of control (maximal BDNF concentration  =  8 nM) values were plotted for the indicated concentrations of each ligand (n = 2 ± SD for each data point). Data indicate activation of TrkB-dependent SH2 recruitment by TrkB mAbs using the TrkB EFC assay.

Figure 3. Functional activity of TrkB mAbs in the primary cortico-striatal neuronal co-culture system.

Rat primary cortico-striatal co-cultures were stimulated with BDNF, mAb 38B8 and mAb 29D7 and profiled by (A) western blot; agonists tested at 10 nM over the indicated incubation times using non-transfected co-cultures; or (B) mHTT-induced co-culture protocol over the indicated concentration range, as described in the Materials and Methods. For western blot, lysates were prepared and run on SDS-PAGE gels and transferred to membranes that were then hybridized to the corresponding primary antibodies, as described in the Methods and Materials. For mHTT-induced co-culture assay, % Rescue (Normalized to in-plate controls; 0.22 nM BDNF (100% Rescue) and vehicle (0% Rescue)) values were plotted for striatal neurons over the indicated concentrations of each ligand (n = 6 ± SD for each data point). Data demonstrates activation of TrkB phosphorylation signal transduction cascade and rescue of mHTT-induced neuronal toxicity by TrkB mAbs.

Figure 4. Functional activity of digested 38B8 mAb in the TrkB NFAT reporter gene assays.

CellSensor® TrkB-NFAT-bla CHO-K1 cells were stimulated with (A) BDNF, mAb 38B8, 38B8 Fab or 38B8 F(ab’)2 over the indicated concentration range for 5 hours (agonist mode) or (B) 38B8 Fab for 1 hour prior to 0.3 nM BDNF stimulation for 4 hours (antagonist mode) before beta-lactamase assay was performed as described in Methods. % of control (maximal BDNF concentration  =  9 nM) values were plotted for the indicated concentrations of each ligand (n = 2 ± SD for each data point).

Figure 5. Literature-reported TrkB small molecule agonists.

Literature-reporteded TrkB small molecule agonists tested were amitriptyline , 7,8-dihydroxyflavone , , N-acetyl serotonin , LM22A-4 , and B_AG (cyclic peptide . Compound structure or sequence as well as corresponding reported properties and activities are indicated.

Figure 6. TrkB activation assays: small molecule pharmacology.

Literature-claimed TrkB small molecule agonists tested in TrkB agonist (A, C) and PAM (B, D) modes in the CellSensor® NFAT reporter gene (A, B) and PathHunter® EFC (C, D) assays; for PAM mode the compounds are tested in the presence of ∼EC₂₀ concentration of BDNF (40 pM, in (B) and 250 pM in (D). For agonist mode, BDNF potency (EC₅₀) was 323 pM (A) and 823 pM (C), thereby validating the assay systems. No agonist or PAM activity was observed for the literature-claimed molecules. Curves n = 2 +/– SD, representative data where all experiments were performed in duplicate or triplicate (% of control based on maximal BDNF response). (E) Western blot assessment of ERK1/2, AKT and TrkB phosphorylation status. The CellSensor® TrkB NFAT-bla CHO-K1 cell line was treated for 5 hours using two concentrations of each test compound, as described above. As a positive control for TrkB, ERK1/2 and AKT phosphorylation BDNF was applied at three concentrations (0.15, 0.53 and 1.15 nM). Experiment also configured with a 1 hour treatment regimen; as with 5 hour data, no compound-mediated phosphorylation effects were observed at this earlier time point (data not shown).

Figure 7. mHTT-induced neuronal toxicity: 7,8-dihydroxyflavone and LM22A-4 pharmacology.

Using our primary rat cortico-striatal co-culture system we stimulated with 7,8-dihydroxyflavone, LM22A-4 or BDNF over the indicated concentration range according to the mHTT-induced co-culture protocol (A), as described in Methods. % Rescue (Normalized to in-plate controls; 1 nM BDNF (100% Rescue) and vehicle (0% Rescue)) values were plotted for striatal neurons over the indicated concentrations of each ligand (n = 6 ± SD for each data point). Rat primary cortico-striatal cells (non-transfected) were stimulated with 7,8-dihydroxyflavone (B) or LM22A-4 (C) using the indicated concentrations for 15 minutes and profiled by western blot. BDNF- (10 nM) mediated TrkB phosphorylation validated the experimental system.