Autophagy Is Rapidly Induced by Salt Stress and Is Required for Salt Tolerance in Arabidopsis

Salinity stress challenges agriculture and food security globally. Upon salt stress, plant growth slows down, nutrients are recycled, osmolytes are produced, and reallocation of Na+ takes place. Since autophagy is a high-throughput degradation pathway that contributes to nutrient remobilization in plants, we explored the involvement of autophagic flux in salt stress response of Arabidopsis with various approaches. Confocal microscopy of GFP-ATG8a in transgenic Arabidopsis showed that autophagosome formation is induced shortly after salt treatment. Immunoblotting of ATG8s and the autophagy receptor NBR1 confirmed that the level of autophagy peaks within 30 min of salt stress, and then settles to a new homeostasis in Arabidopsis. Such an induction is absent in mutants defective in autophagy. Within 3 h of salt treatment, accumulation of oxidized proteins is alleviated in the wild-type; however, such a reduction is not seen in atg2 or atg7. Consistently, the Arabidopsis atg mutants are hypersensitive to both salt and osmotic stresses, and plants overexpressing ATG8 perform better than the wild-type in germination assays. Quantification of compatible osmolytes further confirmed that the autophagic flux contributes to salt stress adaptation. Imaging of intracellular Na+ revealed that autophagy is required for Na+ sequestration in the central vacuole of root cortex cells following salt treatment. These data suggest that rapid protein turnover through autophagy is a prerequisite for salt stress tolerance in Arabidopsis.


Supplementary Figure 1 Illustration of materials used in this study.
Null mutants for five of the core autophagy machinery genes-atg5, atg7, atg10, atg2, and atg9-are used along with the transgenic Arabidopsis over-expressing GmATG8c (ATG8-OX) for physiological, biochemical, and cell biology studies. ATG5, ATG7 and ATG10 are required for conjugation of ATG8 to phosphatidylethanolamine (PE). ATG5 also recruits ATG8 to the expanding phagophore. ATG2 mediates ATG9 cycling between an unknown membrane source and the phagophore. The incidence of autophagy is analyzed with immunoblotting of the autophagosome marker ATG8 and the autophagy receptor/adaptor NBR1, which binds both poly-ubiquitinated cargo and ATG8 in selective autophagy. To monitor autophagic flux, the vacuolar lumen alkalizer Concanamycin A (ConA) is used to preserve autophagic bodies in the lytic vacuole.

Supplementary Figure 3 The anti-GmATG8c antisera preferentially detect the non-lipidated ATG8s.
Ten-day-old vertically grown seedlings of wild-type and atg7 were collected for protein extraction. Total membrane fraction (100,000 g, 1h) were collected, and the solubilized samples either treated or not with phospholipase D (PLD) were analyzed with immunoblotting. CE, crude seedling extract prior to fractionation; S, soluble fraction obtained from 100,000 g centrifugation; Mem, membrane fraction obtained from centrifugation and solubilized in Triton X-100. Fractions were treated with PLD at 37 ℃ for 1 h before separated on SDS-PAGE containing 6 M urea. Very faint bands representing ATG8-PE were detected in the WT samples without PLD treatment, but not in PLD-treated WT or in atg7. These low Mw bands (marked with *) likely represent ATG8-PE.

Supplementary Figure 4 Autophagy is not induced by salt treatment in autophagy mutants.
The level of autophagy is represented by comparing the ATG8 protein levels between Concanamycin A (ConA)-treated and untreated samples at the same time points. The level of selective autophagy is represented by differences in the amount of NBR1 in ConA+/-samples. All SDS-PAGE gels contained 6 M urea. Anti-Tubulin antisera were used as internal control. Only a mild reduction in NBR1 was observed in atg9 at 0.5 h of NaCl treatment.

Supplementary Figure 5 Germination of atg mutants, WT, and ATG8-OX on mannitol.
For germination assay on mannitol, stratified seeds were sown on 1/2 MS medium with or without mannitol. Germination rates were scored at a 4-hour interval during first 3 days, then daily until day 7. Four biological replicates were done (n > 200 each). Bar = standard error. *, **, and *** indicate p<0.05 and p<0.01, p<0.001, respectively. Red and black stars indicate hypersensitivity and reduced sensitivity compared with the WT.
Supplementary Figure 6 Root bending assay of atg mutants, WT, and ATG8-OX on 150 mM NaCl.
For root bending assay on NaCl, four-day-old vertically grown seedlings were transferred from 1/2 MS medium to 1/2 MS medium supplemented with 150 mM NaCl. To protect the cotyledons from bleaching, a sterilized plastic strip (8 mm wide) was placed on top of the medium, and the individual seedlings were placed with their cotyledons touching the strip only. The plates were then inverted. After 2 days, the plates were scanned. On 150 mM NaCl, insufficient root-bending was observed in atg5 and atg7.

Supplementary Figure 7 CoroNa Green staining of root tips of in ATG8-OX lines.
(A) Five-day-old three ATG8-OX lines seedlings incubated in liquid 1/2 MS containing 100 mM NaCl for 6 hours, for 6 hours plus 2 hours of CoroNa Green AM (5 μM) (green) staining were scanned with a SP5 (Leica, Germany) confocal microscope as described (Meier et al., 2006;Oh et al., 2010). Plasma membrane was stained with FM4-64 (5 μM) (red) right before scanning. Confocal settings were completely same as for NaCl-treated seedlings shown in the manuscript. Bar = 25 µm. (B) Quantified Fluorescent intensity unit (FIU) of (A).
Supplementary Figure 8 CoroNa Green staining of root tips without NaCl.