Effects of Mutations on the Aggregation Propensity of the Human Prion-Like Protein hnRNPA2B1

Abstract

Hundreds of human proteins contain prion-like domains, which are a subset of low-complexity domains with high amino acid compositional similarity to yeast prion domains. A recently characterized mutation in the prion-like domain of the human heterogeneous nuclear ribonucleoprotein hnRNPA2B1 increases the aggregation propensity of the protein and causes multisystem proteinopathy. The mutant protein forms cytoplasmic inclusions when expressed in Drosophila, the mutation accelerates aggregation in vitro, and the mutant prion-like domain can substitute for a portion of a yeast prion domain in supporting prion activity. To examine the relationship between amino acid sequence and aggregation propensity, we made a diverse set of point mutations in the hnRNPA2B1 prion-like domain. We found that the effects on prion formation in Saccharomyces cerevisiae and aggregation in vitro could be predicted entirely based on amino acid composition. However, composition was an imperfect predictor of inclusion formation in Drosophila; while most mutations showed similar behaviors in yeast, in vitro, and in Drosophila, a few showed anomalous behavior. Collectively, these results demonstrate the significant progress that has been made in predicting the effects of mutations on intrinsic aggregation propensity while also highlighting the challenges of predicting the effects of mutations in more complex organisms.

Keywords: aggregation; low-complexity domain; prion-like; prions.

FIG 1

hnRNPA2 contains a predicted prion-like domain. (A) Schematic of the hnRNPA2 domain architecture. (B) The disease-associated D290V mutation increases predicted prion propensity. PAPA scores (green) and FoldIndex scores (red) were calculated for hnRNPA2 wild type (WT) and D290V. The dashed line indicates a PAPA score of 0.05, the threshold that was most effective at separating prion-like domains with and without prion activity (23). Regions with high PAPA scores and negative FoldIndex scores are predicted to be prion prone. Adapted from the work of Kim et al. (18).

FIG 2

PAPA accurately predicts prion-promoting mutations at the 290 position. (A) Schematic of wild-type Sup35 and the hnRNPA2-Sup35 chimeric protein (18). Sup35 contains three domains: an N-terminal prion domain, a highly charged middle (M) domain, and a C-terminal domain that is responsible for Sup35's translation termination function. The prion domain contains two parts: a nucleation domain (ND) that is required for prion formation and an oligopeptide repeat domain (ORD) that is dispensable for prion nucleation but is required for prion propagation. In the hnRNPA2-Sup35 fusion, the ND (amino acids 3 to 40) of Sup35 was replaced with the core PrLD (amino acids 261 to 303) from hnRNPA2B1. (B) Western blot analysis of endogenous expression of full-length wild-type (WT) and mutant hnRNPA2-Sup35 chimeric proteins, using an antibody to the Sup35 C-terminal domain. (C) Western blot analysis of overexpression of PrLD-GFP fusions, using an antibody to GFP. The NM domain of each hnRNPA2-Sup35 chimera was fused to GFP and expressed from the GAL1 promoter. (D) Effects of different amino acids at the D290 position. [psi⁻] strains were generated that expressed hnRNPA2-Sup35 fusion proteins with the indicated substitution at the D290 position as the sole copy of Sup35 in the cell. The strains were transformed with either an empty vector (endogenous expression) or a plasmid expressing the matching PrLD-GFP mutant under the control of the GAL1 promoter (PrLD overexpression). Cells were grown in galactose dropout medium for 3 days, and then 10-fold serial dilutions were plated onto medium lacking adenine to select for [PSI⁺] and medium containing adenine to test for cell viability. PAPA scores for each amino acid are indicated. The wild type (D290) and D290V were previously reported (18). (E) Quantification of Ade⁺ colony formation. Serial dilutions of the galactose cultures from panel D were plated onto full plates containing medium with and without adenine. The frequency of Ade⁺ colony formation was determined as the ratio of colonies formed with and without adenine. Data represent means ± SDs (n ≥ 3). (F) Curability of Ade⁺ colonies. For each mutant, eight individual Ade⁺ isolates were grown on YPD (−) or YPD plus 4 mM guanidine HCl (+). Cells were then restreaked onto YPD to test for loss of the Ade⁺ phenotype. (G) The Ade⁺ phenotype is associated with protein aggregation. For the indicated mutants, Ade⁻ and Ade⁺ cells were transformed with a plasmid expressing the matching PrLD-GFP mutant under the control of the GAL1 promoter. Cells were grown for 1 h in galactose dropout medium and visualized by confocal microscopy.

FIG 3

Additive mutations. (A) The indicated hnRNPA2-Sup35 mutants were tested for prion formation. D276V enhances Ade⁺ colony formation, albeit less than the D290V mutation. The D276V D290V double mutant shows substantially higher levels of Ade⁺ colony formation than the D290V mutant alone, even forming Ade⁺ colonies in the absence of PrLD overexpression. (B) Quantification of Ade⁺ colony formation. Data represent means ± SDs (n ≥ 3). (C) Western blot analysis of endogenous expression of full-length hnRNPA2-Sup35 chimeric proteins and overexpression of PrLD-GFP fusions.

FIG 4

Compensatory mutations. (A) Prion-inhibiting mutations effectively offset the effects of the D290V mutation. Y283 in the hnRNPA2-Sup35 (D290V) fusion was replaced with either other prion-promoting amino acids (F, I, and V), a neutral amino acid (N), or prion-inhibiting amino acids (R, E, P, D, and K). Each of the predicted prion-inhibiting amino acids partially or completely reversed the effects of the D290V mutation. (B) Quantification of Ade⁺ colony formation. Data represent means ± SDs (n ≥ 3). (C) Y288 in the hnRNPA2-Sup35 (D290V) fusion was replaced with either a prion-inhibiting proline or a prion-promoting isoleucine. (D) Western blot analysis of endogenous expression of full-length hnRNPA2-Sup35 chimeric proteins and overexpression of PrLD-GFP fusions.

FIG 5

Predicted strong steric zipper segments are neither necessary nor sufficient for prion activity. (A and B) ZipperDB analysis (24) of the core PrLDs of hnRNPA2 wild type and D290V mutant. Segments with a Rosetta energy below −23.0 kcal/mol are predicted to form steric zippers. The D290V mutation creates a strong predicted steric zipper segment from amino acids 287 to 292. The remainder of the core PrLD is not scored by ZipperDB due to the presence of prolines at positions 262 and 298. Adapted from the work of Kim et al. (18). (C) The Y283K mutation blocks prion formation by the hnRNPA2-Sup35 (D290V) mutant (Fig. 4) but does not affect the predicted strong steric zipper segment. (D) Both ΔD290 and a ΔD276 ΔD290 double mutant substantially increase Ade⁺ colony formation by the hnRNPA2-Sup35 fusion. (E) Quantification of Ade⁺ colony formation. Data represent means ± SDs (n ≥ 3). (F) Western blot analysis of endogenous expression of full-length hnRNPA2-Sup35 chimeric proteins and overexpression of PrLD-GFP fusions. (G and H) Neither the ΔD290 nor ΔD276 ΔD290 mutations are predicted to create a strong steric zipper.

FIG 6

Effects of mutations in Drosophila. (A) Adult fly thoraces were stained with anti-hnRNPA2B1 (green), Texas Red-X phalloidin (red), and DAPI (blue). Wild-type hnRNPA2 localizes exclusively to the nuclei, whereas the D290V mutant also forms cytoplasmic foci. The other mutants show a range of localization patterns, including much more substantial cytoplasmic foci for the D276V D290V double mutant. Examples of cytoplasmic and nuclear foci are indicated with arrows and arrowheads, respectively. (B) Thoraces of adult flies were dissected and sequential extractions were performed to examine the solubility profile of hnRNPA2. (C) Quantification of the blot shown in panel B. Data represent means ± SEMs (n = 3). ****, P < 0.0001 (two-way analysis of variance [ANOVA] with Bonferroni's post hoc test. ns, not significant.

FIG 7

In vitro amyloid formation by hnRNPA2 mutants. (A and B) Synthetic 35-amino-acid peptides from the hnRNPA2 core PrLD were generated with the indicated mutations. Peptides were resuspended under denaturing conditions and then diluted to initiate amyloid formation. Reaction mixtures were incubated at room temperature with intermittent shaking. Amyloid formation was monitored by thioflavin T fluorescence. Data represent means ± SEMs, with error bars shown for every 10th data point; n = 3. (C) Electron micrographs of amyloid formation assays after 48 h. Scale bars, 500 nm.