Figure 1.
αB-crystallin protects ARPE-19 cells from MGO-induced apoptosis.
ARPE-19 cells treated with 1 mM MGO for 4 h were harvested 48 h post-treatment (scRNA, scrambled siRNA; siRNA, RNAi against αB-crystallin; Ctrl, control). (A) Viability assay. Neither 1 mM MGO alone nor 1 mM MGO together with the scrambled siRNA control reduced the viability of ARPE-19 cells. However, silencing of αB-crystallin sensitized ARPE-19 cells and reduced their viability in the presence of 1 mM MGO. ** P<0.01. (B) DNA electrophoresis and PFGE. Whereas 1 mM MGO in conjunction with siRNA against αB-crystallin did not produce ladder-like DNA fragments on a conventional agarose gel (left), PFGE revealed the disintegration of nuclear DNA into giant fragments of 1–2 Mbp and high molecular-weight fragments of 100–1000 Kbp in length cells (right). (C) Nuclear morphology revealed by Hoechst staining. Silencing of αB-crystallin induced nuclear condensation of ARPE-19 cells in response to MGO. (D) Representative histograms indicating cell cycle progression and induction of apoptosis (Apo, the percentage of the population undergoing apoptosis). Silencing αB-crystallin induced the accumulation of subdiploid apoptotic ARPE-19 cells in response to MGO. (E) A western blot of apoptosis-related proteins. Silencing of αB-crystallin induced the degradation of procaspase-3 and -7 and PARP as well as the formation of their cleavage products following MGO treatment. In addition, downregulation of survivin, XIAP and Bcl-2 were observed (β-actin was used as a loading control). (F) Flow cytometry results indicating mitochondrial membrane potential (MMP). Silencing αB-crystallin induced a reduction in MMP in ARPE-19 cells in response to MGO. (G) Confocal microscopy images showing the subcellular localization of cytochrome c. Silencing αB-crystallin induced the release of cytochrome c from ARPE-19 mitochondria following MGO treatment. (PI, propidium iodide; Cyto C, cytochrome c).
Figure 2.
MGO modulates subcellular localization and phosphorylation of αB-crystallin at three different serine residues.
ARPE-19 cells were treated with 2 mM MGO. (A) A western blot of αB-crystallin. Although MGO slightly decreased the expression of αB-crystallin, MGO downregulated the phosphorylation of αB-crystallin at Ser-19 and Ser-45. Conversely, MGO upregulated the phosphorylation of αB-crystallin at Ser-59 (β-actin was used as a loading control). (B) A western blot indicating the fractionation of cells treated with MGO. Cytoplasmic αB-crystallin was suppressed by MGO treatment, but nuclear αB-crystallin remained. P-αB-crystallin-Ser-19 was found in both the cytoplasm and nuclei of untreated control cells. MGO treatment caused the loss of cytoplasmic P-αB-crystallin-Ser-19, whereas its nuclear distribution remained unchanged. P-αB-crystallin-Ser-45 was found in the nuclei but not in the cytoplasm of untreated control cells, and its nuclear distribution remained following MGO treatment. P-αB-crystallin-Ser-59 was found primarily in the cytoplasm of untreated control cells, whereas MGO treatment increased the amount of nuclear P-αB-crystallin-Ser-59 while inducing its loss from the cytoplasm. (C–F) Confocal microscopy images showing the subcellular location of αB-crystallin and phosphorylation of αB-crystallin on Ser-19, Ser-45 or Ser-59. SC-35 and αB-crystallin were stained with Texas Red and FITC, respectively. (C) Confocal microscopy images showing αB-crystallin. αB-crystallin was localized to the cytoplasm and nuclei of untreated control ARPE-19 cells. Within nuclei, αB-crystallin was localized to SC35 speckles. MGO treatment substantially reduced cytoplasmic αB-crystallin, but the distribution of αB-crystallin in SC35 speckles was largely unaffected. (D) Confocal microscopy images of P-αB-crystallin-Ser-19. P-αB-crystallin-Ser-19 was found both in the cytoplasm and within nuclear SC35 speckles in untreated control cells. MGO treatment reduced cytoplasmic P-αB-crystallin-Ser-19, but its localization within SC35 speckles was unchanged. (E) Confocal microscopy images of P-αB-crystallin-Ser-45. P-αB-crystallin-Ser-45 was found in nuclear SC35 speckles in untreated control cells, and its distribution was unaffected by MGO treatment. (F) Confocal microscopy images of P-αB-crystallin-Ser-59. P-αB-crystallin-Ser-59 was found primarily in the cytoplasm of untreated control cells. MGO treatment reduced cytoplasmic P-αB-crystallin-Ser-59, causing its relocalization to the perinuclear region.
Figure 3.
MGO-induced apoptosis is mediated by caspase-2 and PIDDosome formation.
ARPE-19 cells were treated with 2 mM MGO. (A) A western blot showing the expression levels of several caspase-2 variants and splicing factors. MGO treatment triggered the production of the cleavage products of caspase-2L and -2S. Importantly, MGO upregulated the proform of caspase-2L, whereas MGO downregulated the proform of caspase-2S. MGO treatment substantially increased the expression SC-35 but slightly decreased the expression of SF2/ASF and hnRNP A1 (β-actin was used as a loading control). (B) An RT-PCR assay of caspase-2L and -2S expression. The expression levels of caspase-2L and -2S were increased by MGO treatment. The expression level of caspase-2L RNA transcripts was markedly increased in a time-dependent manner. (C) A western blot showing the expression of PIDDosome components. MGO treatment downregulated full-length PIDD but did not significantly alter levels of RAIDD expression. Importantly, MGO treatment upregulated PIDD-C and cause the production of PIDD-CC products (β-actin was used as a loading control). (D) Co-immunoprecipitation experiments demonstrating PIDDosome formation. Interaction between caspase-2L and RAIDD was observed following MGO treatment, whereas no interaction occurred between these proteins in untreated control cells. The interaction between PIDD and RAIDD was increased in cells treated with MGO. (E) A co-immunoprecipitation assay showing the interaction between αB-crystallin and RAIDD or PIDD. In the control ARPE-19 cells, αB-crystallin bound to RAIDD and PIDD, but these interactions were abolished by MGO treatment.
Figure 4.
Interactions between αB-crystallin and various caspase subtypes.
(A) A co-immunoprecipitation assay demonstrating the association between αB-crystallin and various caspase subtypes. All of the caspase subtypes tested interacted with αB-crystallin in the untreated control ARPE-19 cells. The interactions between αB-crystallin and each caspase subtype were reduced by MGO treatment. (B) A co-immunoprecipitation assay illustrating the interaction between αB-crystallin and caspase-2L; fractionated cell lysates (nuclear and cytoplasmic fractions) were used. Caspase-2L bound to αB-crystallin in the cytoplasmic fraction but not in the nuclear fraction of the untreated control ARPE-19 cells. The interaction between caspase-2L and un-phosphorylated αB-crystallin was abolished by treatment with MGO; this was evident at 48 h post-treatment. P-αB-crystallin-Ser19 and -Ser45 did not interact with caspase-2L in the untreated control cells or the MGO-treated cells. However, the interaction between P-αB-crystallin-Ser59 and caspase-2L was enhanced by MGO treatment in both fractions. (C) A co-immunoprecipitation assay indicating the interaction between αB-crystallin and caspase-7. Caspase-7 bound to unphosphorylated αB-crystallin in both the cytoplasm and nuclei of untreated control cells, but this interaction was decreased in both fractions when the cells were treated with MGO. P-αB-crystallin-Ser19 did not interact with caspase-7 in the untreated control cells or in MGO-treated cells in either fraction. P-αB-crystallin-Ser45 and -Ser59 interacted with caspase-7 in the nucleus, but not in the cytoplasm, and this interaction was found to be sustained or even increased following MGO treatment. (D) A co-immunoprecipitation assay indicating the interaction between αB-crystallin and caspase-3. Caspase-3 bound to unphosphorylated αB-crystallin both in the cytoplasm and the nucleus of untreated control cells, but this interaction was decreased in both fractions following MGO treatment. P-αB-crystallin-Ser19 did not interact with caspase-3 in untreated control cells or in MGO-treated cells. P-αB-crystallin-Ser45 interacted with caspase-3 in the nucleus, but not in the cytoplasm, and this interaction was found to be sustained or even increased following MGO treatment. P-αB-crystallin-Ser59 did not interact with caspase-3 in the untreated control cells in either fractions; however, an interaction was induced by MGO treatment, with a maximal effect observed at 24 h in both fractions. (E-G) Confocal microscopy images showing the association of caspase subtypes with SC35. The fluorescence intensity profiles of SC35 and each caspase subtype are depicted. (E) Caspase-2L is not located in the nuclei of ARPE-19 cells. (F) Caspase-7 is dispersed throughout the nuclei of the control cells. MGO treatment induced the localization of caspase-7 into SC35 speckles. (G) Caspase-3 is dispersed in the nuclei of control cells. MGO treatment induced the localization of caspase-3 into SC35 speckles.
Figure 5.
αB-crystallin inhibits the activation of caspase-2 and -7 in the nuclei of control ARPE-19 cells.
ARPE-19 cells were transfected with constructs containing the wild-type αB-crystallin gene (CRYAB) or non-phosphorylatable cryab mutants (S19A, S45A or S59A). Following transfection, cells were treated with 2 mM MGO for 24 h, harvested and then assayed. (A) Confocal microscopy images showing that replacing residues Ser19, Ser45 or Ser59 with alanine prevented the nuclear localization of αB-crystallin in SC35 speckles. (B) A western blot indicating the localization of the nonphosphorylatable αB-crystallin mutants S19A, S45A and S59A. Because the αB-crystallin gene was fused to Flag sequences in the wild-type and mutant constructs, an anti-Flag antibody (Flag) was used to detect both wild-type and mutant proteins. Mutant αB-crystallin proteins were mostly retained in the cytoplasmic fraction (N, nuclear fraction; C, cytoplasmic fraction. β-actin and histone H3 were used as loading controls). (C) A viability assay. The introduction of αB-crystallin mutants did not affect the viability of cells. (D) An RT-PCR assay indicating that transfection of these mutants induced caspase-2L cleavage without concomitant caspase-2L upregulation. (E) A western blot. Transfection of the αB-crystallin mutants induced the formation of cleavage products from caspase-2L and -7, but not from caspase-3 or PARP. Transfection of the αB-crystallin mutants did not affect PIDD-CC and RAIDD formation (Flag, an anti-Flag antibody was used to detect the wild type and mutant αB-crystallin transfected into the cells). (F) A western blot showing the cleavage products of various caspase subtypes. Wild-type and mutant αB-crystallin constructs were transfected into cells and both nuclear and cytoplasmic fractions were isolated from the transfected cells. Cleaved caspase-7 was found in the both the cytoplasmic and nuclear compartments, whereas the cleavage products of caspase-2L appeared only in the cytoplasm. No cleavage product of caspase-3 was observed in either the cytoplasm or the nuclei. (G) A western blot indicating the effects of silencing αB-crystallin. siRNA against αB-crystallin efficiently reduced αB-crystallin expression, caused the production of a caspase-2L cleavage product, and induced the degradation of procaspase-7. However, no changes were observed in the expression or production of cleavage products for caspase-3, PARP or PIDD-CC (β-actin was used as a loading control). (H) A western blot indicating the effects of silencing caspase-7. Caspase-7 siRNA prevented the activation of caspase-2L in S19A, S45A or S59A mutant-expressing ARPE19 cells. Caspase-7 siRNA used in this experiment efficiently reduced the expression of caspase-7.
Figure 6.
Phosphorylation of αB-crystallin at serines 19, 45 and 59 plays a pivotal role in preventing MGO-induced ARPE19 apoptosis.
ARPE-19 cells were transfected with constructs containing wild-type or nonphosphorylatable αB-crystallin mutants (S19A, S45A or S59A). (A) A co-immunoprecipitation assay indicating the interactions between αB-crystallin and caspase-3 or -7 as well as between PIDD and RAIDD in mutant cells. While the interactions between αB-crystallin and caspase-3 or -7 were decreased, the interaction between RAIDD and PIDD was increased in mutant-expressing cells (Flag, Flag antibody used to detect the wild type and mutant αB-crystallin). (B) The viability of mutant cells treated with 2 mM MGO for 24 h. All mutant-expressing cells showed a significant reduction in viability at 24 h post-MGO treatment. ** P<0.01. (C) A western blot indicating the effects of mutant αB-crystallin expression in response to MGO treatment. MGO upregulated the expression of caspase-2L in mutant-expressing cells. The cleavage products of caspase-2L, -7 and -3, PARP, and PIDD-CC were also increased in the mutant cells (Flag, Flag antibody used to detect the transfected wild-type and mutant αB-crystallin). (D) RT-PCR assays showing the expression of caspase-2L in mutant-expressing cells. Caspase-2L mRNA transcripts were increased in the mutant cells treated with MGO.