Resolution of Two Steps in Botulinum Neurotoxin Serotype A1 Light Chain Localization to the Intracellular Plasma Membrane

Abstract

Botulinum neurotoxin serotype A (BoNT/A) is the most potent protein toxin to humans. BoNT/A light chain (LC/A) cleavage of the membrane-bound SNAP-25 has been well-characterized, but how LC/A traffics to the plasma membrane to target SNAP-25 is unknown. Of the eight BoNT/A subtypes (A1-A8), LC/A3 has a unique short duration of action and low potency that correlate to the intracellular steady state of LC/A, where LC/A1 is associated with the plasma membrane and LC/A3 is present in the cytosol. Steady-state and live imaging of LC/A3-A1 chimeras identified a two-step process where the LC/A N terminus bound intracellular vesicles, which facilitated an internal α-helical-rich domain to mediate LC/A plasma membrane association. The propensity of LC/A variants for membrane association correlated with enhanced BoNT/A potency. Understanding the basis for light chain intracellular localization provides insight to mechanisms underlying BoNT/A potency, which can be extended to applications as a human therapy.

Keywords: SNAP-25; botulinum neurotoxin; botulinum neurotoxin serotype A; cellular microbiology; subtype; toxins.

Figure 2

Blastp alignment of A3 versus A1. The primary amino acid sequences of A3LM (top) ACA57525, A1 (middle) ACS66881, and A2 (bottom) ADB85243 were analyzed by Blastp [32]. The bottom line depicts identical amino acids between A1, A2, and A3LM (_*); conserved amino acids (:); and non-conserved amino acids ( ). Shown are the N region (amino acids 1–17, red) and the low homology domain (LHD, green), which comprises amino acids 268–357 and has ~60% identity between A3LM and A1. Cysteine residues are bolded and underlined.

Figure 1

(A) Structure of botulinum neurotoxin A (BoNT/A) (PDB: 3BTA). (B) Structural alignments of A1 and A3LM. The light chain (LC) is shown in gray, the N in red, and LHD in green A1 (PDB: 1XTG) and middle A3LM (PDB: 7DVL) [31] and right A1 and A3LM merged with N (residues 1–17, red) and LHD (268–357, green) highlighted. Merge of A1 and A3LM was prepared with PyMol software. Circle schematic of LC:N (red), LHD (green), regions outside (gray), HC:HCN (blue), and HCC (magenta).

Figure 3

Intracellular localization of A3 N sequence variants. (Left) After overnight transfections, N2A cells were fixed with 4% paraformaldehyde and imaged for EGFP fluorescence (excitation 488 nm, emission 509 nm). Representative images show the steady-state localization of EGFP, EGFP-LC/A3 Loch Maree (A3LM), EGFP-LC/A3V (A3V), EGFP-LC/A3LM(R¹¹A) (A3LM(R¹¹A)), and EGFP-LC/A1 (A1). (Right) Percentage of EGFP membrane-bound or present in the cytosol. Ten random fields were selected and counted for membrane (Upper) or cytosolic (Lower) localization. Mean and SEM were evaluated, with ordinary one-way ANOVA with Dunnett’s multiple comparisons test using A3 LM as the control column: ns, not significant; * p < 0.01; ** p < 0.05; *** p < 0.001.

Figure 4

LC/A1 N and LHD are necessary and sufficient to localize A3V to the plasma membrane of N2A cells. After overnight transfection, N2A cells were fixed with 4% paraformaldehyde and imaged for EGFP fluorescence (excitation 488 nm and emission 509 nm). (Left) Schematic representation of the construct with the region of interest depicted as EGFP fluorescent tag, N terminus residues 1–17 (N), and low homology domain (LHD) residues 268–357. Representative images show the steady-state localization of EGFP, GFP-LC/A3V, EGFP-LC/A3V(A1-N), EGFP-LC/A3V(A1-LHD), EGFP- LC/A3V(A1-N, LHD), EGFP-LC/A1(A3LM-LHD), or EGFP-LC/A1 (A1). (Right) Percentage of EGFP membrane- or cytosol-localized. Ten random fields were selected and counted for membrane (upper) or cytosolic (lower) localization. Mean and SEM were evaluated, with ordinary one-way ANOVA with Dunnett’s multiple comparisons test using A1 as the control column: ns, not significant; *** p < 0.001.

Figure 5

Time-lapse imaging of EGFP-LC/A. After a five-hour transfection of the indicated EGFP-LC/A (A); EGFP-LC/A1 (A1) (B) EGFP-LC/A3V (A3V) (C); EGFP-LC/A3V(A1-N), (A3V(A1-N)) (D); and EGFP-LC/A3V(A1-LHD) (A3V(A1-LHD)) (E). EGFP-LC/A3V(A1-N, LHD) (A3V(A1-N, LHD)) and N2As were imaged on a Nikon Eclipse Ti2 microscope equipped with a W1 Spinning Disc, Orca Flash CMOS camera, and 60× oil-immersion objective (CFI Plan Apo λ, 1.4 NA objective) confocal microscope [33]. Live-cell images were obtained every 10 s for 10 min. Images were deconvoluted with Nikon Elements Deconvolution Software (Version 6).

Figure 6

Structures of the proposed A1 and A3(LM) intramolecular interactions. (Top) Crystal structure of A1 (PDB:1XTG) and (Bottom) A3LM (PDB:7DVL) distance between glutamine (Q⁷) and lysine (K⁸⁹) as measured with PyMOL. The N of A1 and A3LM is highlighted in red with the remaining regions of LC in green.

Figure 7

Model for the anterograde trafficking and localization of A1 to the plasma membrane. The N (red) of A1 associates with the extracellular surface of the vesicle; the LHD (green) allows A1 to efficiently associate with the plasma membrane. LHD transitions both soluble and vesicular-bound A1 to the plasma membrane independent of the N of A1, but transfers are more efficient when LC is vesicle-associated.