1. Xba I 19 7.7 6.2 4.3 kbp Sac I Hind III Fig. S1 Southern blot analysis using genomic DNA from Nona Bokra plants. Five µg of Nona Bokra genomic DNA was digested with XbaI, SacI, or HindIII. The full length of OsHKT2;1 cDNA from cultivar Nipponbare was labelled by 32 P and used as probe. High stringency hybridization (65 ºC) is performed as described previously (Horie et al. 2001)

2. a b Fig. S2 Structural features of No-OsHKT2;2/1 from a salt tolerant landrace Nona Bokra and its selectivity pore forming regions. a A schematic drawing of the No-OsHKT2;2/1 protein in comparison with those of the OsHKT2;1 and Po-OsHKT2;2. Single upper letter with a line indicates the 88th amino acid in each protein (S, serine; G, glycine), which corresponds to the highly conserved amino acid position (in general G) in the first p-loop domain of HKT transporters. Boxed regions in light grey and dark grey correspond to amino acid sequences of OsHKT2;1 and Po-OsHKT2;2, respectively. A black region in No-OsHKT2;2/1 represents amino acids identical to both OsHKT2. b Alignments of four p-loop domains (P A to P D ) of representative class II HKT transporters from rice (OsHKTs), wheat (TahKT2;1) and barley (HvHKT2;1). A cartoon represents transmembrane-pore-transmembrane (MPM) topology including a p-loop domain. Arrowheads indicate the position of highly conserved glycine residue in each p-loop domain

3. Vector No-OsHKT2;2/1 EGFP-No-OsHKT2;2/1 1.0K 0.1K Fig. S3 Complementation test using high-affinity K + uptake deficient mutant (strain CY162) of S. cerevisiae. CY162 cells harboring either vector (pYES2) or pYES2::No-OsHKT2;2/1 or pYES2::EGFP-No-OsHKT2;2/1 were grown on AP medium containing 1.0 mM (left) or 0.1 mM (right) KCl. 1:10 serial dilutions of each transformant were spotted on the plates with the starting OD 600 of 0.1. All plates were incubated at 30 ºC for 4d and pictures were taken afterwards