Misfolding of transthyretin in vivo is controlled by the redox environment and macromolecular crowding

Abstract

Transthyretin (TTR) amyloidosis is a progressive disorder characterized by peripheral neuropathy, autonomic dysfunction, and cardiomyopathy. The precise mechanism by which TTR misfolds and forms fibrils in vivo remains incompletely understood, posing challenges to the development of effective therapeutics. In this study, we reveal that the recently identified nonnative pathological species of TTR (NNTTR), which is enriched in the plasma of ttr-val30met gene carriers, exhibits strong amyloidogenic properties, making it a promising therapeutic target. Notably, we demonstrate that NNTTR formation is dependent on an intermolecular disulfide bond and can be promoted by oxidative conditions while being effectively suppressed by reducing agents. The formation of this disulfide bond is incompatible with the native TTR fold, thereby necessitating structural flexibility. We further show that this required flexibility can be constrained using tetramer-stabilizing drugs, thereby suppressing NNTTR formation. Interestingly, the flexibility is also hindered by macromolecular crowding, and NNTTR formation is strongly suppressed by the high protein concentration in plasma. This suppression is released upon dilution, which thus promotes NNTTR formation in areas with lower protein content, highlighting a potential link to the interstitial space, brain, and vitreous body of the eye, where TTR-amyloid is frequently observed. In summary, we demonstrate that NNTTR displays strong amyloidogenic features, underscoring its potential as a therapeutic target. We identify the redox environment and macromolecular crowding as key modulatory factors. Our findings propose a mechanistic explanation for TTR misfolding and suggest a novel therapeutic approach.

Keywords: amyloid; cysteine; disulfide; macromolecular crowding; redox; transthyretin.

Figure 1

NNTTR from ttr-val30met carriers contains a high proportion of a nonnative disulfide bond.A, analysis of the levels of NNTTR in plasma from ten heterozygote carriers of ttr-val30met and 10 controls with the ttr wt gene, conducted using a sandwich-based ELISA with MAB_39-44. All plasmas were diluted 400X before analysis. The bars show sex (female (F) and male (M)) and age, as well as standard deviation. B, SDS-PAGE and Western blot analysis after immunoprecipitation of plasma from five ttr-val30met carriers. The immunoprecipitated material was analyzed after boiling in SDS-loading buffer under nonreducing versus reducing conditions, followed by SDS-PAGE and Western blotting. Detection of TTR was performed using the polyclonal rabbit anti-TTR antibody TTR_49-127 according to the standard procedures. The sex (female (F) and male (M)) and age of the individuals are indicated in the figure. rTTR-V30M boiled in SDS-loading buffer under reducing conditions was included as a control. NNTTR, nonnative variants of TTR; rTTR, recombinant TTR; TTR, transthyretin.

Figure 2

Disulfide formation between subunits of TTR correlates with formation of NNTTR.A, the presence of NNTTR was assessed using the MAB_39-44 sandwich-based ELISA on rTTR-V30M (4 μM in PBS), incubated at 37 °C for 0 to 2 h with either 40 μM diamide or 1 mM BME. To exclude potential nonspecific effects from diamide, the rTTR-C10S, V30M mutant, lacking cysteines, was incubated with 40 μM diamide under identical conditions. All proteins were monitored using nonreducing SDS-PAGE stained for proteins. B, SDS-PAGE and Western blot analysis were performed on rTTR-V30M, incubated overnight at 37 °C in PBS with 40 μM diamide. Lanes 1 and 2 show the total sample, before immunoprecipitation, boiled in nonreducing versus reducing SDS-loading, respectively. Lanes 3 and 4 show the analysis of the material captured by immunoprecipitation with MAB_39-44. TTR was detected by Western blotting using the polyclonal rabbit anti-TTR_49-127 antibody, following standard procedures. BME, β-mercaptoethanol; NNTTR, nonnative variants of TTR; rTTR, recombinant TTR; TTR, transthyretin.

Figure 3

Isolation of disulfide-linked TTR-V30M and AFM analysis.A, SDS-PAGE stained for protein using Coomassie brilliant blue, before and after isolation of disulfide-linked rTTR-V30M using anion-exchange chromatography. B, AFM analysis of fully reduced TTR-V30M after incubation in PBS, pH 7.4, at 55 °C for 72 h (scale bar represents 500 nm). C, AFM analysis of disulfide-linked TTR-V30M after incubation in PBS, pH 7.4, at 55 °C for 72 h (scale bar represents 500 nm). AFM, atomic force microscopy; rTTR, recombinant TTR; TTR, transthyretin.

Figure 4

The disulfide-linked NNTTR retains a tetrameric structure but acquires a high sensitivity for proteolytic digestion by trypsin.A, SEC analysis of rTTR-V30M having a fraction of NNTTR corresponding to around 70%, (black line), and fully reduced native rTTR-V30M (red line). The inset shows the fractions of the NNTTR after SEC analyzed on SDS-PAGE verifying the presence of a high fraction of disulfide-linked dimers. B, nonreducing SDS-PAGE analysis of rTTR-V30M containing 25% of NNTTR and 75% native tetramers after exposure to various concentrations of bovine trypsin in PBS for 1 h at 37 °C. NNTTR, nonnative variants of TTR; SEC, size-exclusion chromatography; rTTR, recombinant TTR; TTR, transthyretin.

Figure 5

The formation of NNTTR is effectively suppressed by reducing agents. The relative formation of NNTTR from rTTR-V30M-Cys at 4 μM incubated in PBS at 37 °C for 24 h, using the MAB_39-44 sandwich-based ELISA. A, shows the relative increase of NNTTR between 0 and 24 h. B–F, show the formation of NNTTR after 24 h as a function of 0 to 125 μM GSH; 0 to 125 μM NAC; 0 to 125 μM BME; 0 to 125 μM TCEP; and 0 to 2500 μM cystine, respectively. All samples were diluted 5000X in PBS-T before analysis and error bars show the standard deviation. All experiments have been repeated at least three times. BME, β-mercaptoethanol; NAC, N-acetylcysteine; NNTTR, nonnative variants of TTR; PBS-T, PBS supplied with 0.3% Tween-20; rTTR, recombinant TTR; TCEP, tris(2-carboxyethyl) phosphine.

Figure 6

Dilution of human plasma in PBS-EDTA strongly promotes NNTTR formation. NNTTR formation was investigated using a sandwich-based ELISA with MAB_39-44 on plasma from 6 ttr-val30met carriers A–F, shown as a function of 1X–32X dilution using PBS-EDTA, or dialyzed against PBS-EDTA using a 3.5 kDa membrane. All samples were incubated for 24 h at 37 °C and normalized to 5000X dilution using PBS-T before analysis. The sex (female (F) and male (M) and age of the individuals are given in each figure. All experiments have been repeated at least 3 times. Error bars indicate the standard deviation. NNTTR, nonnative variants of TTR; PBS-T, PBS supplied with 0.3% Tween-20.

Figure 7

The formation of NNTTR is effectively suppressed by macromolecular crowding and tetramer-stabilizing ligands. Plasma from a ttr-val30met carrier was incubated for 24 h after 20X dilution in PBS-EDTA, pH 7.4 along with (A) PEG 6000 in concentrations of 0 to 30%, (B) Ficoll 70 in concentrations of 0 to 30%, and (C) HEWL in concentrations of 0 to 100 mg/ml. A corresponding analysis was performed on 4 μM of rTTR-V30M-Cys incubated along with (D) PEG 6000 in concentrations of 0 to 30% (E) Ficoll 70 in concentrations of 0 to 30%, and (F) HEWL in concentrations of 0 to 100 mg/ml. To probe the effect of tetramer stabilizing agents plasma from a ttr-val30met carrier was incubated for 24 h after 20X dilution in PBS-EDTA, pH 7.4 along with (G) tafamidis 0 to 10 μM. H, diflunisal 0 to 10 μM. I, luteolin 0 to 10 μM. All samples were diluted 5000X using PBS-T before analysis using a sandwich-based ELISA with MAB_39-44. Error bars show standard deviations, and all experiments have been repeated at least 3 times. HEWL, hen egg-white lysozyme; PBS-T, PBS supplied with 0.3% Tween-20; NNTTR, nonnative variants of TTR; rTTR, recombinant TTR.

Figure 8

Conversion of native TTR into NNTTR through dilution is effectively suppressed by NAC. The propensity to form NNTTR was investigated in plasma from 3 ttr-val30met carriers, as a function of various concentrations of NAC. All samples were normalized to 5000X dilution using PBS-T before analysis for NNTTR using the MAB_39-44 sandwich ELISA. All bars show standard deviations and all experiments have been repeated at least three times. NAC, N-acetylcysteine; NNTTR, nonnative variants of TTR; PBS-T, PBS supplied with 0.3% Tween-20; TTR, transthyretin.

Figure 9

Model of the native tetrameric assembly of TTR WT based on the crystal structure F141.pdb. TTR is shown from two different views where the single cysteine located in position 10 on each subunit has been highlighted in yellow. The closest distance between two adjacent cysteines is around 25 Å. TTR, transthyretin.

Figure 10

A simplified schematic illustration showing how native TTR is stabilized by the highly crowded environment of human plasma. The diffusion of TTR from the vascular system into the interstitial space lowers this stabilizing effect and increases the formation rate of NNTTR. The figure illustrates how the process can be suppressed by both reducing agents as well as stabilizing ligands. The novel epitopes exposed on NNTTR are susceptible to proteolytic digest and although both full-length TTR as well as fragments after proteolytic digest can form amyloid the latter cannot reverse to its native state. NNTTR, nonnative variants of TTR; TTR, transthyretin.