α-Gal A missense variants associated with Fabry disease can lead to ER stress and induction of the unfolded protein response

Abstract

Anderson-Fabry Disease (FD) is an X-linked lysosomal disorder caused by mutations in GLA, the gene encoding the lysosomal hydrolase α-galactosidase A (α-Gal A), leading to accumulation of glycosphingolipids in the lysosomes. FD is a multisystemic disorder leading to progressive cardiovascular, cerebrovascular and kidney dysfunction. Phenotypes are divided in two main classes, classic or non-classic, depending on substrate accumulation, age at onset, disease manifestation, severity and progression. The more severe classical phenotype is generally associated with mutations leading to absent or strongly reduced α-Gal A activity, while mutations with higher residual activity generally lead to the non-classical one. Approximately 70% of the over 1,000 Fabry disease-associated mutations are missense mutations, some leading to endoplasmic reticulum (ER) retention of mutant protein. We hypothesized that such mutations could be associated, besides the well-known absence of α-Gal A function/activity, to a possible gain of function effect due to production of a misfolded protein. We hence expressed α-Gal A missense mutations in HEK293 GLA ^-/- cells and investigated the localization of mutant protein and induction of ER stress and of the unfolded protein response (UPR). We selected a panel of 7 missense mutations, including mutants shown to have residual or no activity in vitro. Immunofluorescence analysis showed that mutants with residual activity have decreased lysosomal localization compared with wild type, and partial retention in the ER, while missense mutants with no residual activity are fully retained in the ER. UPR (ATF6 branch) was significantly induced by all but two mutants, with clear correlation with the extent of ER retention and the predicted mutation structural effect. These data identify a new molecular pathway, associated with gain of function effect, possibly involved in pathogenesis of FD.

Keywords: ER stress; Fabry disease; Missense mutations; Unfolded protein response; α-galactosidase A.

Fig. 1

Expression of α-Gal A isoforms in transfected HEK293 GLA^−/−cells. (A) Representative Western blot showing the absence of α-Gal A in Crispr/Cas9 edited cells. GAPDH was used as loading control. (B) Schematic representation of human α-Gal A. The grey box represents the leader sequence; the light blue trident represent glycosylation sites, represent active sites, the yellow box represents the substrate binding site. The position of selected variants is shown. In yellow, variants with residual activity; in red, variants with no residual activity. The residual activity of D231N variant is reported 0 or 0.5% of wild type activity [5,9]. (C) α-Gal A expression in transiently transfected HEK293 GLA^−/−cells assessed by real-time RT-qPCR. Expression is normalized to HPRT1. Data are expressed as mean ± s.d. (n = 3 independent experiments). (D) Western blot analysis showing α-Gal A expression in lysates of HEK293 GLA^−/− cells transfected with different α-Gal A isoforms. Actin is shown as a loading control.

Fig. 2

Intracellular localization of wild type and mutant α-Gal A isoforms. (A) Representative immunofluorescence images showing the cellular distribution of wild type or mutant α-Gal A (red) and of the lysosomal protein Lamp1 (green) in transiently transfected HEK293 GLA^−/− cells. (B) Quantification of α-Gal A (GLA) signal co-localizing with the lysosomal marker Lamp1 (n = 20 cells). Data are shown as vertical scatterplots indicating mean ± s.d. (all mutant isoforms were compared to wild type, one-way ANOVA followed by Dunnett's multiple comparison test). *** p < 0.001.

Fig. 3

ER retention of missense α-Gal A mutants. (A) Representative immunofluorescence images showing the cellular distribution of wild type or mutant α-Gal A isoforms (red) and of the ER marker KDEL (green) in transiently transfected HEK293 GLA^−/− cells. (B) Quantification of α-Gal A (GLA) signal co-localizing with the ER marker KDEL (n = 20 cells). Data are shown as vertical scatterplots indicating mean ± s.d. (all mutant isoforms were compared to wild type, one-way ANOVA followed by Dunnett's multiple comparison test). **p < 0.01; *** p < 0.001.

Fig. 4

UPR induction in cells expressing missense α-Gal A mutations. (A-C) Real-time RT-qPCR showing expression levels of HSPA5, XBP1s and DDIT3, normalized to HPRT1. Data are expressed as mean ± s.d. (n = 3 independent experiments) relative to wild type cells. (D) Western blot analysis of PERK in HEK293 GLA^−/− cells transfected with different α-Gal A isoforms. T indicates incubation with tunicamycin (2 μg/ml for 14 h), as positive control of PERK phosphorylation (molecular weight shift). (E) ATF6 activation assessed through the use of a luciferase-based, ATF6 reporter construct. Data are expressed as mean ± s.d. (n = 6 independent experiments). (A-E) All mutant isoforms were compared to wild type, one-way ANOVA followed by Dunnett's multiple comparison test. **p < 0.01, ***p < 0.001.

Fig. 5

Predicted structural effect of Fabry disease missense mutations. (A) Histogram of root mean squared deviation (RMSD) values of α‑carbon atoms of the investigated missense mutants as a read out of their structural alteration, according to fabry-database.org. (B) Scatter plot of RMSD values and % of α-Gal A (GLA) in ER values of the selected missense mutations. (C) Scatter plot of RMSD values and ATF6 reporter values of the selected missense mutations. (D) Histogram of RMSD values of n = 663 missense mutations according to fabry-database.org. The horizontal dashed red line indicates 0.05 RMSD threshold.

Fig. 6

Proposed model of UPR induction, mutation structural impact and GLA residual activity. In wild type condition, α-Gal A enters the secretory pathway to reach its lysosomal localization. At steady state the protein is almost completely localized in the lysosome and active. Misfolded mutants reported to have residual activity and low structural impact are partially ER retained inducing ER stress and UPR to an extent that is inversely correlated with ER retention and residual enzymatic activity. Mutants shown to have high structural impact and no residual activity are fully retained in the ER and induce robust ER stress and UPR.