Osmolytes and crowders regulate aggregation of the cancer-related L106R mutant of the Axin protein

Abstract

Protein aggregation is involved in a variety of diseases, including neurodegenerative diseases and cancer. The cellular environment is crowded by a plethora of cosolutes comprising small molecules and biomacromolecules at high concentrations, which may influence the aggregation of proteins in vivo. To account for the effect of cosolutes on cancer-related protein aggregation, we studied their effect on the aggregation of the cancer-related L106R mutant of the Axin protein. Axin is a key player in the Wnt signaling pathway, and the L106R mutation in its RGS domain results in a native molten globule that tends to form native-like aggregates. This results in uncontrolled activation of the Wnt signaling pathway, leading to cancer. We monitored the aggregation process of Axin RGS L106R in vitro in the presence of a wide ensemble of cosolutes including polyols, amino acids, betaine, and polyethylene glycol crowders. Except myo-inositol, all polyols decreased RGS L106R aggregation, with carbohydrates exerting the strongest inhibition. Conversely, betaine and polyethylene glycols enhanced aggregation. These results are consistent with the reported effects of osmolytes and crowders on the stability of molten globular proteins and with both amorphous and amyloid aggregation mechanisms. We suggest a model of Axin L106R aggregation in vivo, whereby molecularly small osmolytes keep the protein as a free soluble molecule but the increased crowding of the bound state by macromolecules induces its aggregation at the nanoscale. To our knowledge, this is the first systematic study on the effect of osmolytes and crowders on a process of native-like aggregation involved in pathology, as it sheds light on the contribution of cosolutes to the onset of cancer as a protein misfolding disease and on the relevance of aggregation in the molecular etiology of cancer.

Figure 1

Biophysical characterization of RGS WT and RGS L106R. (A) CD spectra of RGS WT (blue) and RGS L106R (red) at 4°C. (B) Melting curves of RGS WT (blue) and RGS L106R (red) recorded by CD at 222 nm. ANS fluorescence spectra in the presence of (C) RGS WT and (D) RGS L106R at 4°C are shown, with RGS/ANS molar ratio varying from 1:1 to 1:12. To see this figure in color, go online.

Figure 2

Biophysical characterization of RGS L106R aggregation. (A) DLS and (B) CD of RGS L106R acquired at fixed time points over 24 h of incubation at 25°C: time 0 (red), 30 min (blue), 1 h (light blue), 4 h (orange), 12 h (yellow), 24 h (black). (C) Time-course kinetics of ANS fluorescence at 470 nm in the presence (red) and in the absence (black) of RGS L106R at 25°C. (D) TEM image of RGS L106R after 12 h of incubation at 25°C. The scale bar represents 100 nm. To see this figure in color, go online.

Figure 3

RGS WT does not display aggregation properties. (A) DLS and (B) CD of RGS WT acquired at fixed time points over 24 h of incubation at 25°C: time 0 (red), 30 min (blue), 1 h (light blue), 4 h (orange), 12 h (yellow), 24 h (black). (C) ANS time-course kinetics in the presence (blue) and in the absence (black) of RGS WT at 25°C. (D) TEM image of RGS WT after 12 h of incubation at 25°C. The scale bar represents 100 nm. To see this figure in color, go online.

Figure 4

Effect of polyols on RGS L106R aggregation. (A) Time-course ANS kinetics of RGS L106R aggregation at 25°C in the absence (black) and in the presence of 750 mM glycerol (purple), erythritol (pink), xylitol (light green), sorbitol (light blue), myo-inositol (brown), glucose (red), sucrose (blue), maltose (orange), and trehalose (green). Each trace represents the average of at least three independent measurements. (B) Dependence of k_ANS of RGS L106R aggregation in the presence of 750 mM glycerol, erythritol, xylitol, glucose, and sucrose (blue line) and of sorbitol, myo-inositol, maltose, and trehalose (red) on the molecular weight of the polyol osmolytes. The error bars represent the standard deviation (SD) of the k_ANS obtained from the fitting of at least three independent measurements. To see this figure in color, go online.

Figure 5

Effect of amino acids on RGS L106R aggregation. (A) Time-course ANS kinetics of RGS L106R aggregation at 25°C in the absence (black) and in the presence of 750 mM glycine (pink), alanine (purple), serine (light green), proline (blue), glycylglycine (light blue), hydroxyproline (red), and betaine (green). Each trace represents the average of at least three independent measurements. (B) Dependence of the k_ANS of RGS L106R aggregation in the presence of 750 mM betaine (orange), glycine, alanine, serine, proline, glycylglycine, and hydroxyproline (green) on the molecular weight of the amino acid osmolytes. The error bars represent the SD on the k_ANS obtained from the fitting of at least three independent measurements. The dashed green line is a guide for the eye. To see this figure in color, go online.

Figure 6

Effect of crowders on RGS L106R aggregation. (A) Time-course ANS kinetics of RGS L106R aggregation at 25°C in the absence (black) and in the presence of PEG 400 (green), PEG 3350 (blue), and PEG 6000 (red) at an ethylene glycol monomer concentration of 750 mM. Each trace represents the average of at least three independent measurements. (B) Dependence of the k_ANS of RGS L106R aggregation in the presence of PEG 400, PEG 3350, and PEG 6000 on the molecular weight of the crowders at an ethylene glycol monomer concentration of 750 mM. The error bars represent the SD on the k_ANS obtained from the fitting of at least three independent measurements. The purple line is a guide for the eye. To see this figure in color, go online.

Figure 7

Apparent rate of RGS L106R aggregation. k_ANS (A) and ANS_end (B) of RGS L106R aggregation in the absence of cosolutes (pink) and in the presence of polyols (green), amino acids derivatives (red), and crowders (blue) at a monomer concentration of 750 mM are shown. The error bars represent the SD on the k_ANS obtained from the fitting (A) or on the ANS_end (B) of at least three independent measurements. The pale pink stripe represents the variation of k_ANS or ANS_end in the absence of cosolutes as a reference of significance for the effect of the cosolutes. To see this figure in color, go online.

Figure 8

Effect of cosolutes on RGS L106R aggregation upon coincubation. Representative TEM images of RGS L106R after 12 h of incubation at 25°C in the presence of (A) sorbitol, (B) glucose, (C) sucrose, (D) maltose, (E) trehalose, (F) serine, (G) proline, (H) hydroxyproline, (I) glycylglycine, (J) betaine, (K) PEG 3350, and (L) PEG 6000 at a monomer concentration of 750 mM are given. The scale bar represents 100 nm.

Figure 9

Effect of cosolutes on RGS L106R stability. CD of RGS L106R incubated at 25°C in the absence (black) or in the presence glucose (pale blue), sucrose (light blue), trehalose (blue), PEG 3350 (orange), and PEG 6000 (red) at a monomer concentration of 750 mM after 12 h of incubation (A) or immediately after mixing (B) is given. CD melting curves of RGS L106R in the absence (black) and in the presence of (C) glucose (pale blue), sucrose (light blue), and trehalose (blue) or (D) PEG 3350 (orange) and PEG 6000 (red) at a monomer concentration of 750 mM are shown. To see this figure in color, go online.

Figure 10

Suggested model of RGS L106R aggregation in vitro. RGS WT has a compact native fold (green) with a cluster of hydrophobic residues buried in the core of the protein (yellow). (A) The L106R cancer-related mutation induces partial unfolding and exposure of the core residues in extended hydrophobic patches on the protein surface. (B) During the first hour of incubation, tetrameric association products of RGS L106R are formed via hydrophobic interactions between the exposed patches. (C) Further aggregation proceeds by piling up of the tetrameric oligomers to form a worm-like, unbranched aggregate. To see this figure in color, go online.

Figure 11

Model of Axin L106R aggregation in vivo. (A) In the cytoplasm, Axin L106R fluctuates as a monomer, with the DIX (blue) and mutated RGS (green) domains free to move. In the presence of osmolytes (red), the hydrophobic patches (yellow) formed on RGS surface upon L106R-induced partial unfolding remain partially hidden from the solvent, thereby inhibiting aggregation. (B) Axin forms the destruction complex with GSK-3β (red), β-catenin (violet), CK1 (purple), and the large, mostly intrinsically disordered scaffold protein APC (light blue) through the RGS domain. The high local concentration of disordered strands within the multiprotein complex triggers the destabilization of mutated RGS and the exposure of the hydrophobic patches. (C) Axin L106R aggregates via mutated RGS within the protein complex, turning it nonfunctional. To see this figure in color, go online.