A Calcium Sensor Homolog Required for Plant Salt Tolerance

See allHide authors and affiliations

Science  19 Jun 1998:
Vol. 280, Issue 5371, pp. 1943-1945
DOI: 10.1126/science.280.5371.1943


Excessive sodium (Na+) in salinized soils inhibits plant growth and development. A mutation in the SOS3 gene renders Arabidopsis thaliana plants hypersensitive to Na+-induced growth inhibition. SOS3 encodes a protein that shares significant sequence similarity with the calcineurin B subunit from yeast and neuronal calcium sensors from animals. The results suggest that intracellular calcium signaling through a calcineurin-like pathway mediates the beneficial effect of calcium on plant salt tolerance.

Soil salinity stresses plant growth and agricultural productivity (1). For many salt-sensitive plants, glycophytes, which include most crop plants, a major part of the growth inhibition is caused by excess Na+(2). High Na+ disrupts potassium (K+) nutrition and, when accumulated in the cytoplasm, inhibits many enzymes (3–5). Adding calcium (Ca2+) to root growth medium enhances salt tolerance in glycophytic plants (6–8). Ca2+ sustains K+ transport and K+-Na+ selectivity in Na+-challenged plants (8).

In Arabidopsis thaliana, the recessivesos3 mutant is hypersensitive to Na+ and another alkali ion, Li+ (9). Under salt stress,sos3 plants accumulate more Na+ and retain less K+ than the wild type. sos3 mutant plants are also incapable of growing under low-K+ conditions. Increased Ca2+ in the culture medium can partially suppress the Na+ hypersensitivity of sos3 plants and completely suppress the defect in K+ nutrition (9). These phenotypes suggest that the SOS3 gene product is part of a crucial pathway for mediating the beneficial effect of Ca2+ during salt stress (9).

The SOS3 locus is on chromosome V between the molecular markers nga139 and CDPK9 (9). Mapping of yeast artificial chromosome (YAC) clones containingnga139 or CDPK9 (or both) (10) placedSOS3 between the left end of YAC EG20H2 (20H2L) and the left end of YAC CIC12F2 (12F2L) (Fig. 1A). Bacterial artificial chromosome (BAC) clones hybridizing to 20H2L or 12F2L (or to both) were isolated, and their end fragments were mapped genetically (Fig. 1A). A binary cosmid clone (COS21-1) hybridizing to the left end of BAC TAMU1E7 (1E7L) and the right end of BAC TAMU2D1 (2D1R) was identified that complements the sos3 mutant (11). 1E7L and 2D1R were sequenced, and the sequences match the MOP9 p1 clone sequenced by the Japanese KAZUSA genome sequencing center (12). Genomic DNA fragments corresponding to putative open reading frames (ORFs) between 1E7L and 2D1R were amplified from sos3 mutant plants and sequenced. The analysis revealed a 9–base pair (bp) deletion in the hypothetical P3 gene. The deletion is consistent with the sos3 mutation having been generated by fast neutron bombardment (9). Genomic DNA covering the predicted P3 ORF was amplified from wild-type plants and cloned into the pBIN19 binary vector to complement the sos3mutant (13). Fifty independent transformants were tested, and all complemented the sos3 phenotype (Fig. 2).

Figure 1

Cloning and sequence analysis of theSOS3 gene. (A) Summary of positional cloning. Molecular markers nga139 and CDPK9 were used as the starting points for identifying the overlapping YAC clones (10). RFLP analysis delimited the SOS3 locus to a 120–kilobase (kb) region between the left ends of EG20H2 and CIC12F2 (10). A BAC contig was assembled within this region (10). The left end of TAMU1E7 was used as a probe to isolate cosmid clones for transformation into sos3 mutant plants. The cosmid COS21-1 rescued sos3 mutant phenotypes (11). Facilitated by the release of the genomic sequence of the entire MOP9 P1 clone (12), the SOS3 gene within the complementing cosmid was identified through sequencing candidate genes from the sos3 mutant plants; cM, centimorgans. (B) Gene structure of SOS3 and position of the sos3 mutation. Positions are relative to the initiation codon. Horizontal arrows indicate the left and right borders of the genomic fragment that complemented sos3mutant phenotypes.

Figure 2

Complementation of sos3by the wild-type SOS3 gene. (A) Wild type. (B) sos3. (C) Transgenicsos3 containing the wild-type SOS3 gene. Four-day-old seedlings grown on MS nutrient medium were transferred to a Murashige-Skoog medium supplemented with 100 mM NaCl. The picture was taken 10 days after the transfer to 100 mM NaCl. The plants were grown upside down for observation of root growth after the root-bending assay (5). The wild-type, sos3, and complementedsos3 plants did not show any difference when grown on MS medium without supplementation of NaCl (27).

The transcribed sequence of the SOS3 gene was determined by sequencing several overlapping cDNAs obtained by library screening and by reverse transcriptase polymerase chain reaction (RT-PCR) (14). The deduced amino acid sequence of the SOS3gene is similar to that of a large number of EF hand calcium-binding proteins (15). The protein SOS3 is predicted to contain three potential Ca2+-binding sites (15) (Fig. 3). The deletion in the sos3mutant occurred in a highly conserved region and likely disables Ca2+ binding of the second putative EF hand (Fig. 3).

Figure 3

Sequence comparison of SOS3 (GenBank accession number AF060553) with yeast CnB subunit (16) and frog NCS (19). Dots indicate gaps introduced to maximize the sequence alignment. Residues identical or similar in at least two of the three sequences are shaded dark or light, respectively. The three predicted EF hands (15) of SOS3 are underlined. Asterisks indicate the basic residues in the Ca2+-binding sites of SOS3 where conserved acidic residues are found in other EF hand proteins (15). The double line marks residues deleted in the sos3-1 allele (21).

The proteins most similar to the SOS3 gene product are the B subunit of calcineurin (CnB) (16) and animal neuronal calcium sensors (NCS) (17–19). Like CnB (16) and NCS (18), SOS3 also contains a putative myristoylation motif [MGXXXS/T(K)] (20,21) at the NH2-terminus. The deduced amino acid sequence of SOS3 shows 27 to 31% identity and 49 to 51% similarity with CnB from various organisms. Although a calcineurin-like protein has been implicated to be involved in several plant processes (22), its biochemical or molecular identity has been elusive. In animal cells, calcineurin plays a key role in diverse cellular functions, including T cell activation, neurotransmission, neutrophil migration, and Na+ homeostasis (23). In the yeast Saccharomyces cerevisiae, calcineurin is required to switch K+ transport from low- to high-affinity mode for improved K+-Na+ selectivity during Na+ stress (24). It is also essential for the transcriptional induction of the ENA1 gene encoding a Na+–adenosine triphosphatase that pumps Na+out of the cell (24). Loss-of-function mutations in CnB confer increased sensitivity of yeast cells to Na+inhibition (24). The functional and sequence similarities between SOS3 and yeast CnB indicate that SOS3 may be part of a plant calcineurin.

With animal NCS, SOS3 shares 30 to 31% identity and 49 to 50% similarity in amino acid sequence. NCS belongs to the recoverin subfamily of EF hand calcium-binding proteins that are expressed mainly in the brain or in photoreceptor cells (17–19). Proteins in this subfamily may function either by stimulating protein phosphatases (17) or by inhibiting protein kinases (25).

Salt stress, like drought, elicits a rapid rise in the cytosolic Ca2+ concentration (7, 26). This rise in Ca2+ presumably initiates a signaling cascade, resulting in plant adaptive responses. The sequence similarity between SOS3 and CnB and NCS suggests that SOS3 responds to the Ca2+ signal by activating a protein phosphatase or inhibiting a protein kinase (or by doing both) that then regulates K+ and Na+transport systems. Although there do not appear to be conspicuous differences between the cytosolic Ca2+ signals elicited by drought and salinity (26), subtle differences in their kinetics and subcellular spatial arrangement could result in drought- or salinity-specific responses. The specific role of SOS3 in the tolerance of the ionic but not the osmotic component of salt stress (9) strongly supports the existence of ionic stress–specific calcium signaling.

For plants, the amount and interactions of three abundant soil cations, Ca2+, K+, and Na+, are essential determinants of potassium nutrition and salt tolerance and therefore greatly affect plant productivity. Our evidence suggests that SOS3 mediates the interaction of K+, Na+, and Ca2+.

  • * To whom correspondence should be addressed. E-mail: jkzhu{at}


View Abstract

Navigate This Article