Irreversible aggregation of protein synthesis machinery after focal brain ischemia

Abstract

Focal brain ischemia leads to a slow type of neuronal death in the penumbra that starts several hours after ischemia and continues to mature for days. During this maturation period, blood flow, cellular ATP and ionic homeostasis are gradually recovered in the penumbral region. In striking contrast, protein synthesis is irreversibly inhibited. This study used a rat focal brain ischemia model to investigate whether or not irreversible translational inhibition is due to abnormal aggregation of translational complex components, i.e. the ribosomes and their associated nascent polypeptides, protein synthesis initiation factors and co-translational chaperones. Under electron microscopy, most rosette-shaped polyribosomes were relatively evenly distributed in the cytoplasm of sham-operated control neurons, but clumped into large abnormal aggregates in penumbral neurons subjected to 2 h of focal ischemia followed by 4 h of reperfusion. The abnormal ribosomal protein aggregation lasted until the onset of delayed neuronal death at 24-48 h of reperfusion after ischemia. Biochemical study further suggested that translational complex components, including small ribosomal subunit protein 6 (S6), large subunit protein 28 (L28), eukaryotic initiation factors 2alpha, 4E and 3eta, and co-translational chaperone heat-shock cognate protein 70 (HSC70) and co-chaperone Hdj1, were all irreversibly clumped into large abnormal protein aggregates after ischemia. Translational complex components were also highly ubiquitinated. This study clearly demonstrates that focal ischemia leads to irreversible aggregation of protein synthesis machinery that contributes to neuronal death after focal brain ischemia.

Fig. 1

Schematic drawing of the four brain regions analyzed in this study. The coronal section was taken from Bregma 0.48. Regions 1–3 are in the ischemic hemisphere and region 4 is in the contralateral side.

Fig. 2

(a, b) Electron micrographs of neuronal soma in the penumbral regions from a sham-operated control rat (a) and a rat subjected to 2 h MCAO followed by 24 h of reperfusion (b). The ER, mitochondria (M), nucleus (N) and ribosomal rosettes (arrows) were normally distributed in neurons from a sham-operated brain. After ischemia, ribosomal rosettes and ER-associated ribosomal studs were either dissociated into single ribosomes (b, arrowheads) or clumped into large aggregates (b, arrows). Scale bar = 0.5 μm. (c, d) Higher magnification of ribosomal aggregation after focal ischemia. Brain sections were from the same tissues as in Fig. 2 (a,b). Ribosomal rosettes (c, arrows) and ER-associated ribosomal studs (c, arrowheads) were distributed normally in a sham-operated control neuron. After ischemia, ribosomes were abnormally clumped into large clusters or aggregates (d, arrows). Some aggregates were associated with mitochondria (M). Scale bar = 0.1 μm.

Fig. 3

(a) Sedimentation, (b) immunoblots and (c, d) optical density analyses of translational complex components in sucrose density gradient fractions. Rats were subjected either to sham surgery or 2 h of MCAO followed by 4 and 24 h of reperfusion. The detergent/salt-insoluble protein aggregate-containing fractions were resolved in 10–55% sucrose gradients by ultracentrifugation. (a) One major UV-260 nm absorbance peak between density fractions 11 and 14 from the sham-control sample was resolved in the density gradient (sham, arrow). This peak shifted to higher density fractions 16–21 at 4 h (4 h, arrow, dashed line) and fractions 27–33 at 24 h of reperfusion (24 h, arrow, solid line), respectively. (b) L28 (kDa, immunoblot between 10 and 25 kDa is shown), S6 (between 25 and 45 kDa), eIF2α (30–55 kDa), eIF3η (between 75 and 150 kDa) and eIF4E (between 15 and 35 kDa) were shifted into higher densities after ischemia (4 and 24 h). (c, d) Changes in optical densities of protein bands of L28, S6, eIF2α, eIF4E and eIF3η were evaluated with Kodak 1D image software. The levels of L28 and S6 are presented as folds of controls. The levels of eIFs were undetectable in controls. Therefore, they are presented as optical densities. Data are expressed as mean ± SD (n = 3); *denotes p < 0.05 between control and experimental conditions, one-way anova followed by Fisher’s PLSD post-hoc test.

Fig. 4

Immunoblot analysis of HSC70, Hdj1, ubi-proteins, HSP70 and HSC/HSP70 interacting protein (HIP) in the density gradient fractions. The UV-260 profiles in the gradient fractions were the same as in Fig. 3a. (b) Relative to sham-operated controls, HSC70 (immunoblot membrane between 55 and 90 kDa is shown), Hdj1 (between 30 and 50 kDa) and ubi-proteins (ubi-, > 90 kDa) were shifted into higher densities after ischemia (4 and 24 h). Inducible HSP70 (between 55 and 90 kDa) was absent from the sham-operated control and 4 h of reperfusion but deposited into protein aggregate-containing fractions at 24 h of reperfusion, whereas HIP (between 45 and 75 kDa) was present in the control and gradually disappeared from the protein aggregate-containing fractions after focal ischemia. (c, d) Changes in protein bands of HSC70, Hdj1, Ubi-proteins, HSP70 and HIP from three different individual rat samples were evaluated with Kodak 1D image software. Data are expressed as mean ± SD (n = 4); *denotes p < 0.05 between control and experimental conditions, one-way anova followed by Fisher’s PLSD post hoc test.

Fig. 5

Confocal microscopic images of the penumbral region stained with antibodies against ubiquitin, eIF3η, ribosomal S6 protein, ribosomal L28 protein and HSC70, respectively. Sections are shown from a sham-operated control rat and from rats subjected to 2 h of MCAO followed by 4 and 24 h of reperfusion. The ubiquitin and eIF3η shows aggregation patterns after ischemia (arrows), whereas S6, L28 and HSC70 immunoreactivities were decreased during the post-ischemic phase.