Fig. 5: SPA1 phosphorylates C-terminus of eIF2α.1 and eIF2α.2 in a light-dependent manner.

a Predicted phosphorylation sites located at C-terminal eIF2α are indicated in red. Tyr, Ser or Thr were then replaced with alanine (A) and aspartic acid (D) to generate a phospho-null form and a phospho-mimicking form of eIF2α, respectively. b A semi-in vivo kinase assay showed that SPA1 protein extracted from TAP-SPA1 overexpression line phosphorylates wild-type (WT) eIF2α but not the phospho-null form (6A/8A) of eIF2α (autoradiogram in upper panel). The plant extracts were derived from 4-day-old TAP-SPA1 seedlings incubated either under dark (D) or illuminated under white light for 4 h (L4h) before sampling. Three biological repeats of data showed the same results. c A semi-in vivo kinase assay showed that WT version of eIF2α has higher phosphorylation intensity in TAP-SPA1 than in spaQ under light condition (autoradiogram in upper panel). The plant extracts were derived from 4-day-old TAP-SPA1 and spaQ incubated either under dark (D) or illuminated with 4 h white light (L4h) before sampling. The GST only was used as negative control. The lower panel shows the protein levels in a Coomassie blue staining gel. Three biological repeats of data showed the same results. d Immunoblots showed phosphorylation of eIF2α under both dark and light condition in Col-0 and spaQ. Four-day-old Col-0 and spaQ etiolated seedlings incubated either under dark (D) or illuminated with 4 h white light (L4h) before sampling, and extracted proteins were then separated on 6.5% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) gels containing 15 μM Phos-tag. The slow-migrating band (*) is the phosphorylated form of eIF2α. Three biological repeats of data showed the same results. Treatment of calf intestinal alkaline phosphatase (CIP, +) and inactivated boiled CIP (-) were shown. Rpt5 proteins were used as loading control.