Targeting botulinum neurotoxin persistence by the ubiquitin-proteasome system

Abstract

Botulinum neurotoxins (BoNTs) are the most potent natural toxins known. The effects of BoNT serotype A (BoNT/A) can last several months, whereas the effects of BoNT serotype E (BoNT/E), which shares the same synaptic target, synaptosomal-associated protein 25 (SNAP25), last only several weeks. The long-lasting effects or persistence of BoNT/A, although desirable for therapeutic applications, presents a challenge for medical treatment of BoNT intoxication. Although the mechanisms for BoNT toxicity are well known, little is known about the mechanisms that govern the persistence of the toxins. We show that the recombinant catalytic light chain (LC) of BoNT/E is ubiquitylated and rapidly degraded in cells. In contrast, BoNT/A LC is considerably more stable. Differential susceptibility of the catalytic LCs to ubiquitin-dependent proteolysis therefore might explain the differential persistence of BoNT serotypes. In this regard we show that TRAF2, a RING finger protein implicated in ubiquitylation, selectively associates with BoNT/E LC and promotes its proteasomal degradation. Given these data, we asked whether BoNT/A LC could be targeted for rapid proteasomal degradation by redirecting it to characterized ubiquitin ligase domains. We describe chimeric SNAP25-based ubiquitin ligases that target BoNT/A LC for degradation, reducing its duration in a cellular model for toxin persistence.

Fig. 1.

Expression and localization of recombinant BoNT catalytic light chains. (A) Schematic of SNAP25 and proteolytic fragments generated from BoNT/A and BoNT/E. (B) Colocalization of BoNT/E and BoNT/A LCs in neuronal cells. (C) Design of a luciferase reporter for BoNT proteolytic activity. (D) Cells were transfected with plasmids encoding luciferase reporter and YFP, YFP-LCE, or YFP-LCA, and luminescence was measured with a luminometer 36 h posttransfection. Data shown are mean ± SEM (n = 3; P < 0.01 for YFP-LCA compared with YFP) in arbitrary relative luminescence units (RLU). (E) Cells were transfected as in 1D, treated with DMSO or 20 μM MG132 overnight, and lysates were probed with indicated antibodies.

Fig. 2.

BoNT/E LC is ubiquitylated and degraded in cells. (A) Cells were transfected with plasmids encoding HA-tagged ubiquitin (Ub) and the indicated YFP plasmids. Cells were treated with MG132 and lysed under denaturing conditions. YFP-fused BoNT LCs were immunoprecipitated with GFP antibody. Ubiquitylated LCs were detected with HA-specific antibody. The membrane was stripped and reprobed with GFP monoclonal antibody. IB, immunoblot; IP, immunoprecipitation. (B) Degradation of YFP-LCA and YFP-LCE in transfected human neuroblastoma SH-SY5Y cells assessed by cycloheximide (CHX) chase. (C) Degradation of YFP-LCA and YFP-LCE in transfected HEK293 cells assessed by cycloheximide chase. (D) SH-SY5Y cells were transfected with YFP-LCA or YFP-LCE, and degradation of YFP-LCs was monitored by ³⁵S pulse-chase. (E) Quantification of ³⁵S pulse-chase experiments evaluating the degradation of YFP-LCA and YFP-LCE in SH-SY5Y cells (n = 3; values are mean ± SD). The percentage of ³⁵S-labeled protein remaining is plotted as a function of chase time. (F) Degradation of YFP-LCA and YFP-LCE in HEK293 cells assessed by ³⁵S pulse-chase. (G) Quantification of ³⁵S pulse-chase experiments evaluating the degradation of YFP-LCA and YFP-LCE in HEK293 cells (n = 3; values are mean ± SD). The percentage of ³⁵S-labeled protein remaining is plotted as a function of chase time.

Fig. 3.

TRAF2 interacts with the BoNT/E LC and promotes its degradation. (A) Cells were transfected with HA-tagged TRAF2 and the indicated plasmids. After 36 h, cells were treated with MG132 and lysed in RIPA buffer. YFP-fused BoNT LCs were immunoprecipitated (IP) with GFP antibody and immunoblotted (IB) with the indicated antibodies. (B) Cells were transfected with the indicated plasmids, and degradation of YFP-LCE was assessed by cycloheximide chase. (C) Cells were transfected with the indicated siRNAs, and degradation of YFP-LCE was assessed by cycloheximide chase.

Fig. 4.

SNAP25-based ubiquitin ligase enhances BoNT/A LC degradation. (A) Design of a SNAP25-based ubiquitin ligase for targeting BoNT/A LC. The mutations (D179K, R198T) render mutant SNAP25 noncleavable (SNC) by BoNT/A and BoNT/E. (B) E3 activity of SNAP25-based RING finger ubiquitin ligase assessed by self-ubiquitylation in vitro. (C) Ubiquitylation of BoNT/A LC by SNAP25-based ubiquitin ligase in vitro. (D) Cells were transfected with YFP-LCA, HA-tagged ubiquitin, and the indicated plasmids. Cells were treated with MG132. YFP-LCA was immunoprecipitated (IP) with GFP antibody, and immunoblotted (IB) with the indicated antibodies. Arrow indicates migration of unmodified YFP-LCA. Expression of the ubiquitin ligases was assessed by immunoblotting total-cell lysates with FLAG antibody. (E) Cells were transfected with YFP-LCA and the indicated plasmids. Degradation of YFP-LCA was assessed by ³⁵S pulse-chase. (F) Quantification of ³⁵S pulse-chase experiments evaluating the effects of SNAP25-based ubiquitin ligases on YFP-LCA degradation (n = 3; values are mean ± SD). The percentage of ³⁵S-labeled protein remaining is plotted as a function of chase time.

Fig. 5.

SNAP25-based ubiquitin ligases reduce the duration of BoNT/A LC in a cellular model of BoNT persistence. (A) Cells were induced with doxycycline for 16 h to express GFP-LCE and GFP-LCA. The persistence of YFP-LCs was monitored following doxycycline withdrawal. (B) Cells were induced with doxycycline to express GFP-LCA and cultured in doxycycline-free medium. After 8 h, cells were transfected with the indicated plasmids, and the stability of GFP-LCA was monitored. Cyan fluorescent protein (CFP) served as indicator of transfection efficiency.