Linking Duplication of aCalcium Sensor to Salt Tolerance...The SALT-OVERLY-SENSITIVE (SOS) pathway in Arabidopsis (Arabidopsis thaliana) functions to prevent the toxic accumulation

Linking Duplication of a Calcium Sensor to Salt Tolerancein Eutrema salsugineum1[OPEN]

Shea M. Monihan, Choong-Hwan Ryu, Courtney A. Magness,2 and Karen S. Schumaker3,4

School of Plant Sciences, University of Arizona, Tucson, Arizona 85721

ORCID IDs: 0000-0003-3575-2723 (S.M.M.); 0000-0001-7159-2419 (C.R.); 0000-0001-7703-3861 (C.A.M.); 0000-0002-1958-0223 (K.S.S.).

The SALT-OVERLY-SENSITIVE (SOS) pathway in Arabidopsis (Arabidopsis thaliana) functions to prevent the toxic accumulationof sodium in the cytosol when plants are grown in salt-affected soils. In this pathway, the CALCINEURIN B-LIKE10 (AtCBL10)calcium sensor interacts with the AtSOS2 kinase to activate the AtSOS1 plasma membrane sodium/proton exchanger. CBL10 hasbeen duplicated in Eutrema (Eutrema salsugineum), a salt-tolerant relative of Arabidopsis. Because Eutrema maintains growth insalt-affected soils that kill most crop plants, the duplication of CBL10 provides a unique opportunity to functionally test theoutcome of gene duplication and its link to plant salt tolerance. In Eutrema, individual down-regulation of the duplicated CBL10genes (EsCBL10a and EsCBL10b) decreased growth in the presence of salt and, in combination, led to an even greater decrease,suggesting that both genes function in response to salt and have distinct functions. Cross-species complementation assaysdemonstrated that EsCBL10b has an enhanced ability to activate the SOS pathway while EsCBL10a has a function notperformed by AtCBL10 or EsCBL10b. Chimeric EsCBL10a/EsCBL10b proteins revealed that the specific functions of theEsCBL10 proteins resulted from changes in the amino terminus. The duplication of CBL10 increased calcium-mediatedsignaling capacity in Eutrema and conferred increased salt tolerance to salt-sensitive Arabidopsis.

A growing challenge to food security is the loss ofarable land due to the accumulation of salt in the soil.The estimated annual global cost of crop loss due to soilsalinity is $27.3 billion (Qadir et al., 2014). The Food andAgricultural Organization estimates that 20% of irri-gated land, which produces over 40% of the world’sfood, is affected by salt and 1% to 2% of irrigated landis lost each year due to the accumulation of salt (FAO,2002). Solutions for sustainable agriculture will requireidentification of genes and mechanisms that contributeto salt tolerance to enable exploitation of natural traitvariation and the use of transgenic technologies.

Salt, particularly sodium chloride (NaCl), negativelyaffects plant growth in several ways. The accumulationof salt in the soil restricts water movement into the plant(osmotic stress) and the accumulation of sodium ions ina plant cell interferes with metabolic processes (ionicstress; Munns and Tester, 2008). Molecular studies in

Arabidopsis (Arabidopsis thaliana) have identified genesimportant for alleviating ionic stress and provided in-sight into the pathways in which these genes function.For example, the Arabidopsis SALT-OVERLY-SENSITIVE (SOS) pathway removes sodium from the cell,preventing its toxic accumulation in the cytosol. In thispathway, accumulation of sodium triggers an influx ofcytosolic calcium. Increased calcium is perceived bytwo calcium sensors, SOS3 (roots) and CALCINEURINB-LIKE10 (CBL10, also known as SOS3-LIKE CAL-CIUM-BINDING PROTEIN8; leaves), which bind toand activate the SOS2 Ser/Thr protein kinase (alsoknown as CBL-INTERACTING PROTEIN KINASE24[CIPK24]; Liu and Zhu, 1998; Halfter et al., 2000; Liuet al., 2000; Kim et al., 2007; Quan et al., 2007). SOS2phosphorylates SOS1, a plasma membrane sodium/proton exchanger, initiating transport of sodium out ofthe cell (Shi et al., 2000; Qiu et al., 2002, 2003; Quinteroet al., 2002).

While advances have been made identifying genesimportant for the growth of Arabidopsis in the presenceof salt, Arabidopsis, like many crop plants, is sensitiveto soil salinity (glycophytes; Munns and Tester, 2008).Genetic variability for plant salt tolerance exists, withsome plants maintaining growth longer in high con-centrations of salt (halophytes). Presently, little isknown about the mechanisms underlying halophyteadaptation to salinity (Munns and Tester, 2008).Eutrema (Eutrema salsugineum; formerly Thellungiellahalophila), a small crucifer native to the seashore sa-line soils of Eastern China, is a halophytic relative ofArabidopsis (Bressan et al., 2001; Inan et al., 2004;Amtmann, 2009). Eutrema is emerging as a model for

1This work was supported by the National Science Foundation,Division of Integrative Organismal Systems (award 1552099 to K.S.S.).

2Present address: BASF, Nunhems USA, Inc., West Sacramento,CA 95605.

3Author for contact: [email protected] author.The author responsible for distribution of materials integral to the

findings presented in this article in accordance with the policy de-scribed in the Instructions for Authors (www.plantphysiol.org) is:Karen S. Schumaker ([email protected]).

S.M.M., C.-H.R., and C.A.M. performed the research; S.M.M. andK.S.S. analyzed the data and wrote the article.

[OPEN]Articles can be viewed without a subscription.www.plantphysiol.org/cgi/doi/10.1104/pp.18.01400

1176 Plant Physiology�, March 2019, Vol. 179, pp. 1176–1192, www.plantphysiol.org � 2019 American Society of Plant Biologists. All Rights Reserved.

Dow

nloaded from https://academ

ic.oup.com/plphys/article/179/3/1176/6116554 by guest on 03 August 2021

http://orcid.org/0000-0003-3575-2723

http://orcid.org/0000-0003-3575-2723

http://orcid.org/0000-0001-7159-2419

http://orcid.org/0000-0001-7159-2419

http://orcid.org/0000-0001-7703-3861

http://orcid.org/0000-0001-7703-3861

http://orcid.org/0000-0002-1958-0223

http://orcid.org/0000-0002-1958-0223

http://orcid.org/0000-0003-3575-2723

http://orcid.org/0000-0001-7159-2419

http://orcid.org/0000-0001-7703-3861

http://orcid.org/0000-0002-1958-0223

http://crossmark.crossref.org/dialog/?doi=10.1104/pp.18.01400&domain=pdf&date_stamp=2019-02-16

http://www.plantphysiol.org

mailto:[email protected]

http://www.plantphysiol.org/cgi/doi/10.1104/pp.18.01400

understanding plant adaptation to soil salinity (Inanet al., 2004; Yang et al., 2013). Eutrema does not re-quire salt for optimal growth but is able to survive inconditions that kill most, if not all, crop plants and itstolerance does not appear to rely on specialized mor-phological structures (Inan et al., 2004; Amtmann,2009).Comparative genomic studies have revealed that

plant gene families are largely conserved across species,indicating that plants amplify and modify a basic set ofgenes rather than developing new species-specific genefamilies (Flagel andWendel, 2009). This tendency to usea basic suite of genes suggested that the SOS pathway,identified in Arabidopsis, likely functions in Eutremawith modifications that contribute to its increased salttolerance. Several studies have compared the activity ofthe SOS1 sodium/proton exchanger in Arabidopsis(AtSOS1) and Eutrema (EsSOS1). When each proteinwas expressed in a salt-sensitive strain of yeast (Sac-charomyces cerevisiae), yeast expressing EsSOS1 grewbetter in the presence of salt than yeast expressingAtSOS1, indicating that EsSOS1 has a greater level ofself-activation than AtSOS1 (Oh et al., 2009; Jarvis et al.,2014). When the entire SOS pathways were recon-structed in this salt-sensitive yeast strain, yeast express-ing the Eutrema SOS pathway (EsSOS3, EsSOS2, andEsSOS1) were more salt tolerant than yeast expressingthe Arabidopsis SOS pathway (AtSOS3, AtSOS2, andAtSOS1; Jarvis et al., 2014). While it was not possible todetermine whether the greater growth of yeastexpressing the Eutrema SOS pathway was due toEsSOS1 alone or enhanced activation of EsSOS1 byEsSOS2 and EsSOS3, this study links the Eutrema SOSpathway to salt tolerance. Based on the importanceof the SOS pathway in salt tolerance and its well-established link to calcium signaling, additional puta-tive Eutrema SOS pathway genes were identified andcharacterized. One, CBL10 (EsCBL10), was found to beduplicated (EsCBL10a and EsCBL10b) and is the focus ofthis study.Gene duplication is a major source of genetic diver-

sity and, consequently, adaptive evolution (Ohno, 1970;Kondrashov, 2012). The maintenance of duplicatedgene pairs in a genome can be indicative of an adaptivebenefit conferred by the paralogous genes (Hahn, 2009;Kondrashov, 2012). Because Eutremamaintains growthin salt-affected soils that kill most crop plants, the du-plication of CBL10 provides a unique opportunity tofunctionally test the outcome of gene duplication andits link to plant salt tolerance. To understand the rolesthe duplicated Eutrema CBL10 genes play in responseto salinity, we monitored their expression patterns andreduced their expression individually and in combina-tion. To determine the extent of divergence in CBL10function after duplication and to uncover elements ofthe signaling pathways in which the proteins function,protein activities were examined using cross-speciescomplementation and yeast two-hybrid assays. Toidentify domains and amino acids responsible for thedivergence in function, chimeric EsCBL10 proteins

were generated. We found that, in Eutrema, down-regulation of either of the duplicated EsCBL10 genesdecreased growth in the presence of salt, suggestingthat both genes function in response to salinity. Down-regulation of both EsCBL10 genes in combination led toan even greater decrease in growth, suggesting thegenes have additive effects or different functions. Basedon cross-species complementation assays, we foundthat EsCBL10b has an enhanced ability to activate theSOS pathway while EsCBL10a has a function not per-formed by Arabidopsis CBL10 (AtCBL10) or EsCBL10b.Four kinases that interact with EsCBL10a were identi-fied, and chimeric proteins revealed that the role ofEsCBL10b in the SOS pathway and the role of EsCBL10ain an alternative pathway were the result of changes inthe N termini of the proteins. Analysis of the EsSOS3calcium sensor revealed that EsCBL10a and EsSOS3have partially overlapping functions but also distinctroles in salt tolerance. The duplication of EsCBL10appears to have increased the calcium-mediated sig-naling capacity in Eutrema and the duplicated genesconferred increased salt tolerance when expressed insalt-sensitive Arabidopsis.

RESULTS

Eutrema Is a Salt-Tolerant Relative of Arabidopsis

To compare salt tolerance in Arabidopsis andEutrema, growth of the two species was monitored inthe absence and presence of salt during vegetative andreproductive development. Vegetative (rosette) growthin Arabidopsis was reduced at all salt concentrationswhile Eutrema maintained rosette growth even in thepresence of 300 mM NaCl (Fig. 1A). During reproduc-tive development, Arabidopsis inflorescence number,height, and seed production were reduced at all saltconcentrations, while only small reductions in thesetraits were found in Eutrema (Fig. 1B). These results, incombination with data from previous studies (Bressanet al., 2001; Inan et al., 2004), demonstrate that Eutremais more salt-tolerant than Arabidopsis during bothvegetative and reproductive development. Subsequentstudies focused on understanding if the duplicatedEsCBL10 genes play a role in this salt tolerance.

The Duplicated EsCBL10 Genes Play a Role in Eutrema’sResponse to Salt

Models describing the retention of duplicated genespredict changes in expression and/or gene function(Flagel and Wendel, 2009; Hahn, 2009; Innan andKondrashov, 2010). To uncover differences betweenEsCBL10a and EsCBL10b, transcript accumulation wasexamined and expression of the genes was reducedindividually and in combination. In Arabidopsis, CBL10is alternatively spliced into two primary transcripts:one encoding a functional calcium-binding protein and

Plant Physiol. Vol. 179, 2019 1177

Duplication of CBL10 in Eutrema salsugineum

Dow



one encoding a nonfunctional protein (cannot comple-ment theAtcbl10mutant) likely due to a premature stopcodon in a retained intron (Supplemental Fig. S1; Quanet al., 2007). Reverse transcription-PCR was used todetermine the expression patterns of the EutremaCBL10 genes and to identify alternatively spliced tran-scripts. Expression of EsCBL10b in Eutrema was mostsimilar to the expression of AtCBL10 in Arabidopsis;both transcripts were found in aerial tissues (Fig. 2).When comparing transcript accumulation in leavesand flowers, the EsCBL10b transcript was present atsimilar levels in both organs whereas the AtCBL10transcript was found predominantly in leaves and onlyweakly in flowers (Fig. 2). EsCBL10a was expressedthroughout Eutrema, with similar transcript levels inroots, leaves, and flowers (Fig. 2). Structures of theEsCBL10 splice variants were determined by cloningand sequencing 10 transcripts from each band; multiplevariants with different splicing patterns were identified

(Supplemental Fig. S1). If translated, the majority ofthese variants would likely encode nonfunctionalproteins due to premature stop codons or removal ofdomains required for binding calcium (SupplementalFig. S1). EsCBL10a and EsCBL10b transcripts whoseexon/intron structures were most similar to AtCBL10and likely encode functional proteins (listed firstin Supplemental Fig. S1) were used in experimentsdescribed below.

To determine if the EsCBL10 duplication contributesto salt tolerance, six artificial microRNAs (amiRNAs)were designed to reduce expression of EsCBL10aand EsCBL10b individually and in combination(Supplemental Fig. S2). Only one of the two EsCBL10aamiRNAs was specific, the other reduced transcriptlevels of both genes. The salt tolerance of nine inde-pendently transformed, single insertion, homozygouslines for each target gene was monitored. Two repre-sentative lines were chosen: one line per amiRNA

Figure 1. Eutrema is salt tolerant during vege-tative and reproductive development. A, Ara-bidopsis and Eutrema seeds were germinatedon and grown in soil for 1 week and thenseedlings were left untreated (Control) or trea-ted with NaCl in 50 mM increments every 3 duntil the indicated, final concentration wasreached. Photographs were taken 3 weeks afterthe start of the treatment. Bar = 1 cm for allimages. B, Arabidopsis and Eutrema were ger-minated on and grown in soil for 3 weeks andthen, when inflorescence development began,were left untreated (Control) or treated withNaCl in 50 mM increments every 3 d until theindicated, final concentration was reached.Photographs were taken 4 weeks after the startof the treatment. Bar = 7 cm for all images.

1178 Plant Physiol. Vol. 179, 2019

Monihan et al.

Dow



http://www.plantphysiol.org/cgi/content/full/pp.18.01400/DC1






construct except for EsCBL10a, for which only oneamiRNA construct was used. Seedling weight and rootlength of seedlings grown on solid medium weremeasured as likely components of salt tolerance (shootbiomass or yield) to allow small growth differencesassociated with down-regulation of one or both genesto be measured. Reduced expression of EsCBL10a andEsCBL10b individually and in combination did not af-fect growth in control conditions but decreased growthin the presence of salt (Fig. 3, A and B). Compared withthe wild type, in amiRNA lines targeting EsCBL10a,transcript accumulation was reduced (56% and 32%),resulting in seedling hypersensitivity to salt (23% and21% decrease in fresh weight; Fig. 3, B and C). InamiRNA lines targeting EsCBL10b, transcript accumu-lationwas reduced (45% and 61%), resulting in seedlinghypersensitivity to salt (23% and 22% decrease in freshweight; Fig. 3, B and C). In amiRNA lines targeting bothgenes, transcript accumulation of EsCBL10a andEsCBL10b was reduced (64% and 36% and 61% and42%, respectively), resulting in even greater hypersen-sitivity to salt (63% and 45% decrease in fresh weight;Fig. 3, B and C).When grown in the presence of salt, lines with re-

duced expression of both EsCBL10a and EsCBL10b hadfewer branch roots (77% and 67% decrease) and shorterprimary roots (43% and 26% decrease) compared withthe wild type (Supplemental Fig. S3, A and B). Theeffect of salt on root development in lines targetingEsCBL10a or EsCBL10b individually was variable(Supplemental Fig. S3, A and B).These results suggest that EsCBL10a and EsCBL10b

function in Eutrema’s response to salt and may haveboth overlapping and distinct functions. To provideadditional insight into the inherent biochemical activi-ties of the duplicated proteins and to determine

whether differences in function have arisen throughalterations in the amino acid sequence, in vivo, cross-species complementation assays in single mutant ge-netic backgrounds were performed.

EsCBL10a and EsCBL10b Share AtCBL10 Function

To determine if the Eutrema CBL10 genes are func-tional orthologs of AtCBL10, the coding sequence ofeach gene was expressed in the Arabidopsis cbl10 mu-tant (Atcbl10) downstream of a constitutive promoter.The Cauliflower mosaic virus 35S (CaMV 35S) promoterwas used to ensure that the coding sequence wasstrongly expressed without the presence of introns,which are necessary for full expression when using thenative promoter and which might lead to alternativesplicing. The salt tolerance of five independentlytransformed, single insertion, homozygous lines wasassessed for each gene. Both EsCBL10a and EsCBL10bcomplemented the Atcbl10 salt-sensitive phenotype inall lines examined; however, the degree of comple-mentation varied relative to the growth of wild-typeArabidopsis. Strongly complementing lines had highertranscript accumulation and fully restored growth towild-type levels (Fig. 4). Weakly complementing lineshad lower transcript accumulation and restored growthto 45% of that in the wild type (Fig. 4). Representativestrong and weak lines for each are shown (Fig. 4). InArabidopsis, loss of CBL10 also results in a fertilizationdefect with unfertilized ovules (Monihan et al., 2016).Both EsCBL10a and EsCBL10b restored fertility inthe Atcbl10 mutant grown in the presence of salt(Supplemental Fig. S4). These results demonstrate thatthere is conservation of CBL10 function in Arabidopsisand Eutrema.

EsCBL10a and EsCBL10b Function in Distinct Pathways inResponse to Salt

In Arabidopsis, CBL10 functions in the SOS pathwayto prevent the toxic build-up of sodium in the cytosol(Quan et al., 2007; Lin et al., 2009). Potential roles forEsCBL10a and EsCBL10b in this pathway were inves-tigated using a salt-sensitive strain of yeast (AXT3K)expressing SOS2 and SOS1 from Arabidopsis orEutrema. In AXT3K cells expressing either pathway,salt tolerance was greatest in those cells expressingEsCBL10b and weakest in those expressing EsCBL10a(Fig. 5, A and B); differences in activation were not dueto differences in gene expression (Fig. 5C). Enhancedactivation of the SOS pathway by EsCBL10b could bedue to stronger interaction with AtSOS2; yeast two-hybrid assays were performed to determine thestrength of this interaction. EsCBL10b and AtCBL10interacted more strongly with AtSOS2 than EsCBL10a(Fig. 6) and interaction was orientation-specific(Supplemental Fig. S5).

Figure 2. Expression of EsCBL10a and EsCBL10b differs. Transcriptaccumulation is shown for SOS pathway genes from Arabidopsis andEutrema in roots (R), leaves (L), and stage 14 flowers (F). ELONGATIONFACTOR 1a (EF1-a), Loading control. One representative image ofthree replicates is shown.



Dow







Because EsCBL10a functions in plant responses tosalt (down-regulation in Eutrema results in increasedsalt sensitivity and expression complements theAtcbl10salt-sensitive phenotype) but only weakly activates theSOS pathway, we determined if EsCBL10a can functionin alternative pathways. Even though AtCBL10 andAtSOS3 both activate the SOS pathway, they do notreciprocally complement the Atcbl10 or Atsos3 salt-sensitive phenotypes, indicating that the two proteinsalso function in distinct pathways (Quan et al., 2007). Todetermine if EsCBL10a and EsCBL10b can function in anArabidopsis SOS3 pathway, the genes were expressedin Atsos3 downstream of the CaMV 35S promoter andthe salt tolerance of five independently transformed,single insertion, homozygous lines was determined foreach gene. EsCBL10a but not EsCBL10b complementedthe Atsos3 salt-sensitive phenotype in all lines exam-ined; two representative lines of each are shown (Fig. 7,

A and B). Expression of EsCBL10a in Atsos3 restoredrosette growth (fresh weight) to wild-type levels andpartially restored root growth (Fig. 7B). Transcript ac-cumulation of EsCBL10a and EsCBL10b was similar,suggesting that the inability of EsCBL10b to comple-ment the mutant was not due to lack of gene expression(Fig. 7C).

The ability of EsCBL10a to complement Atsos3 incombination with its weaker interaction with AtSOS2suggested that EsCBL10a may function with a differentkinase. In Arabidopsis, SOS2 belongs to the 26-memberCIPK family, raising the possibility that EsCBL10amight interact with another CIPK to complement theAtsos3 salt-sensitive phenotype. Yeast two-hybridassays were performed with all 26 CIPK proteins toidentify a kinase that interacts specifically withEsCBL10a but not with EsCBL10b or AtCBL10. FourCIPK kinases (AtCIPK13, AtCIPK16, AtCIPK6, and

Figure 3. Reduced expression of EsCBL10a andEsCBL10b results in seedling hypersensitivity tosalt. AmiRNAs were designed to reduce ex-pression of EsCBL10a (E10a) and EsCBL10b(E10b) individually (a1 and b1 and b2, respec-tively) and in combination (ab1 and ab2) andgrowth in the absence (Control) and presence ofsalt (200 mM NaCl) was monitored. WT, Wildtype; -, no amiRNA construct. A, Photographsof wild type and amiRNA lines. Bar = 1 cm forall images. B, Fresh weight was measured toquantify growth. Data are means6 SE of at least30 seedlings per genotype grown in three in-dependent experiments. C, Reverse transcrip-tion quantitative PCR. mRNA levels werenormalized to EsACTIN2 (DCT) and the foldchange in expression relative to the wild typewas calculated (22DDCT). Data are means 6 SE

of three biological replicateswith two technicalreplicates each. For all graphs, different lettersindicate significant differences between geno-types (two-way ANOVA, Tukey-Kramer hon-estly significant difference [HSD], P # 0.05).


Monihan et al.

Dow



AtCIPK18) specifically interacted with EsCBL10a(Fig. 8). In the opposite orientation, interaction was notspecific (Supplemental Fig. S6). Taken together, theseresults suggest that interaction with different proteinpartners might underlie the differences in EsCBL10aand EsCBL10b function.

The EsCBL10a- and EsCBL10b-Specific Functions Reside inthe Amino Terminus

Identification of amino acids or domains that un-derlie specific functions is critical for understandinghow duplication of a gene may contribute to plant ad-aptation. Because the same promoter was used for allEsCBL10a and EsCBL10b constructs, their distinctfunctions likely arose through alterations in biochemi-cal functions via changes in their amino acid sequences.In AtCBL10, the hydrophobic domain has been linkedto membrane localization, the four EF-hand domainsbind calcium, and a Ser in the carboxy terminus of theprotein can be phosphorylated to strengthen interactionwith AtSOS2 (Supplemental Fig. S7; Quan et al., 2007;Lin et al., 2009). Based on crystal structures, it has beensuggested that CBL proteins interact with CIPK pro-teins through the three-dimensional arrangement of theEF-hand domains, which form a hydrophobic pocket to

interact with the CIPK-FISL/NAF domain (a 24-aminoacid CIPK-specific domain; Guo et al., 2001; Sánchez-Barrena et al., 2005; Akaboshi et al., 2008). Alignment ofthe EsCBL10 and AtCBL10 proteins identified two ad-ditional regions in the amino terminus that may beimportant for function: an insertion of seven aminoacids in EsCBL10a and a variable region in whichthe amino acid sequence is distinct in all three CBL10proteins (Supplemental Fig. S7). To identify aminoacids or domains that underlie the distinct functions ofthe EsCBL10 proteins, portions of EsCBL10a andEsCBL10b were exchanged to generate three sets ofchimeric proteins (Supplemental Fig. S7). The first chi-meric proteins fused the amino-terminal half of one ofthe EsCBL10 proteins with the carboxy-terminal halfof the other (Supplemental Fig. S7). The second andthird sets of chimeric proteins progressively narroweddown the region underlying the specific functions(Supplemental Fig. S7). To verify that the chimericproteins are functional, each protein was expressed inthe Atcbl10 mutant downstream of the CaMV 35S pro-moter and the salt tolerance of five independentlytransformed, single insertion, homozygous lines wasdetermined; all chimeric proteins complemented theAtcbl10 salt-sensitive phenotype (Supplemental Figs. S8and S9). To identify regions important for activation ofthe SOS pathway, the chimeric proteins were expressed

Figure 4. EsCBL10a and EsCBL10b complement the Atcbl10 salt-sensitive phenotype. EsCBL10a and EsCBL10bwere expressedin Atcbl10 and growth in the absence (Control) or presence of salt (125 mM NaCl) was monitored. A, Photographs of wild type(WT), Atcbl10, and Atcbl10 expressingAtCBL10 (A10), EsCBL10a (E10a), or EsCBL10b (E10b). Bar = 1 cm for all images. B, Freshweight and the length of the primary root were measured to quantify growth. Data are means 6 SE of at least 24 seedlings pergenotype grown in three independent experiments. For all graphs, different letters indicate significant differences betweengenotypes (two-way ANOVA, Tukey-Kramer HSD, P # 0.05). C, Transcript accumulation. ACTIN, Loading control. One rep-resentative image of three replicates is shown.



Dow











in AXT3K along with AtSOS2 and AtSOS1 and growthin the presence of salt was assessed as an indication ofpathway activity. Based on these experiments, theability of EsCBL10b to strongly activate the SOS path-way resides in the first eight amino acids of the protein(Fig. 9). To identify regions important for complemen-tation ofAtsos3, the chimeric proteins were expressed inthe Atsos3 mutant downstream of the CaMV 35S pro-moter and the salt tolerance of five independentlytransformed, single insertion, homozygous lines wasdetermined (Supplemental Fig. S8C). All chimericproteins with the EsCBL10a hydrophobic domain (aN1,aN2, and bN3) complemented the Atsos3 salt-sensitivephenotype like the full-length protein (major element;

Fig. 10). Like EsCBL10a, the chimeric protein bN2,which has the EsCBL10b hydrophobic domain and theEsCBL10a variable region, complemented Atsos3 ro-sette growth (fresh weight) but not root length (minorelement; Fig. 10).

EsCBL10a and EsSOS3 Function in Distinct Pathways inResponse to Salt

The ability of EsCBL10a to complement the Atsos3salt-sensitive phenotype suggests that two genes inEutrema (EsCBL10a and the AtSOS3 ortholog EsSOS3)may have similar functions. To compare the activities ofthe two proteins, the salt tolerance of Atsos3 seedlingsexpressing EsCBL10a or EsSOS3 was examined. EsC-BL10a and EsSOS3were found to complementAtsos3 inan organ-specific manner; EsCBL10a restored rosettegrowth (fresh weight) while EsSOS3 restored primaryroot growth to wild-type levels (Fig. 7, A and B). Theability of EsSOS3 and EsCBL10a to activate the SOSpathway was compared by expressing the genes inAXT3K with SOS2 and SOS1 from either Arabidopsisor Eutrema. EsSOS3 activated the Arabidopsis andEutrema SOS pathways better than EsCBL10a but notas well as EsCBL10b (Supplemental Fig. S10). To de-termine the extent of EsSOS3’s role in salt tolerance, itwas expressed in the Atcbl10 mutant and the salt tol-erance of five independently transformed, single in-sertion, homozygous lines was determined. WhileEsSOS3 did not restore growth to wild-type or EsC-BL10a levels, seedlings expressing EsSOS3 had longerroots and greater fresh weight than Atcbl10 seedlings(Fig. 4; Supplemental Fig. S11). The difference in theability of EsCBL10a and EsSOS3 to complement Atsos3and Atcbl10 and to activate the SOS pathway suggeststhat, while both play a role in response to salt, thesecalcium sensors have diverged in function.

The Eutrema Calcium Sensors Confer Salt Toleranceto Arabidopsis

Down-regulation of EsCBL10a and EsCBL10b, indi-vidually and in combination, decreased growth ofEutrema, indicating that both genes are necessary forsalt tolerance. To determine if the duplicated genes aresufficient to confer salt tolerance, EsCBL10a andEsCBL10b were expressed in combination in Atcbl10,and EsCBL10a and EsSOS3 in combination in Atsos3.For these studies, the mutants were used instead ofwild-type Arabidopsis to maintain gene compositionconsistent with what is found in Eutrema. Plantsexpressing two Eutrema calcium sensors in combina-tion were generated by crossing homozygous plantsexpressing each gene individually and the salt toleranceof heterozygous seedlings was analyzed.

For plants expressing EsCBL10a and EsCBL10b incombination, two crosses were generated usingstrongly complementing lines and two were generated

Figure 5. EsCBL10b strongly activates the Arabidopsis and EutremaSOS pathways. A and B, Yeast (AXT3K) was transformed with SOS1 andSOS2 fromArabidopsis (A) or Eutrema (B) in combinationwithAtCBL10(A10), EsCBL10a (E10a), or EsCBL10b (E10b). Serial decimal dilutions ofyeast cells were spotted onto control media or media containing125 mM NaCl (A) or 250 mM NaCl (B). Two independently transformedcolonies were assayed in three biological replicates; one representativeimage is shown. C, Transcript accumulation. 18S, Loading control. Onerepresentative image of three replicates is shown.


Monihan et al.

Dow






using weakly complementing lines; a representativecross generated from strong lines is shown (Fig. 11). Themajority of the crosses (three of four) were unable toconfer salt tolerance above what was found in wild-

type Arabidopsis or Atcbl10 expressing AtCBL10. Inthe fourth cross, expression of both genes improvedgrowth relative to wild-type Arabidopsis in the pres-ence of salt (growth increased by 10%; Fig. 11, A and B).

Figure 6. EsCBL10b and AtCBL10 interact more strongly with AtSOS2 than EsCBL10a. AtCBL10 (A10), EsCBL10a (E10a),EsCBL10b (E10b), and AtSOS3 (A3) were fused to the GAL4 DNA-activation domain (AD) and interaction with AtSOS2 fused tothe GAL4 DNA-binding domain (BD) was assessed using yeast two-hybrid assays. Serial decimal dilutions of diploid yeastharboring both constructs were spotted onto synthetic defined media (SD) minus Leu (L) and Trp (W), minus LW and His (H),minus LWH and adenine (A), or with the addition of 0.5 mM 3-amino-1,2,4-triazole (3AT). Two independently mated colonieswere assayed in two biological replicates; one representative image is shown.

Figure 7. EsCBL10a but not EsCBL10b complements the Atsos3 salt-sensitive phenotype. EsCBL10a, EsCBL10b, and EsSOS3were expressed in Atsos3 and growth in the absence (Control) and presence of salt (75 mM NaCl) was monitored. A, Photographsof thewild type (WT),Atsos3, andAtsos3 expressingAtCBL10 (A10), EsCBL10a (E10a), EsCBL10b (E10b),AtSOS3 (A3), or EsSOS3(E3). Bar = 1 cm for all images. B, Fresh weight and the length of the primary root were measured to quantify growth. Data aremeans6 SE of at least 24 seedlings per genotype grown in three independent experiments. For all graphs, different letters indicatesignificant differences between genotypes (two-way ANOVA, Tukey-KramerHSD, P# 0.05). C, Transcript accumulation.ACTIN,Loading control. One representative image of three replicates is shown.



Dow



These results suggest that the ability of EsCBL10a andEsCBL10b to confer salt tolerance is variable; however,coexpression can increase growth in the presence of saltabove that of wild-type Arabidopsis.

In all four crosses expressing EsCBL10a and EsSOS3in combination, seedlings had longer primary roots. A

representative cross in which root length increased by15% relative to the wild type in the presence of salt isshown (Fig. 12, A and B). Transcript accumulation ofEsCBL10a and EsSOS3 was similar in the single anddouble lines (Fig. 12C). These results suggest thatEsCBL10a and EsSOS3 are sufficient to confer salt tol-erance and increase growth in the presence of salt abovethat of wild-type Arabidopsis.

DISCUSSION

The Duplication of the Eutrema CBL10 Calcium SensorIncreased the Complexity of Calcium-Mediated Signalingand Confers Salt Tolerance

Duplication of genes has been shown to underlieplant adaptation by expanding signaling responsesduring growth in adverse environments. For example,gene duplication has been shown to contribute toadaptation to cooler temperatures, tolerance to heavy-metal-polluted soils, and expansion of defense re-sponses (Hanikenne et al., 2008; Gonzales-Vigil et al.,2011; Ilyas et al., 2015; Simon et al., 2015). The dupli-cation of CBL10 in Eutrema provided the opportunityto determine if increasing the number and expandingthe activities of a calcium sensor confer an adaptivebenefit during plant growth in saline soils.

To determine the contribution of EsCBL10a andEsCBL10b to calcium-mediated signaling in response tosalt, the effects of down-regulating their expression

Figure 8. Four Arabidopsis CIPKs interact with EsCBL10a. AtCBL10(A10), EsCBL10a (E10a), EsCBL10b (E10b), and AtSOS3 (A3) were fusedto the GAL4 DNA-activation domain (AD) and interaction with theArabidopsis CIPK proteins fused to the GAL4 DNA-binding domain(BD) was assessed using yeast two-hybrid assays. Serial decimal dilu-tions of diploid yeast harboring both constructs were spotted ontosynthetic defined media (SD) minus Leu (L) and Trp (W) or minus LWand His (H). Two independently mated colonies were assayed in twobiological replicates; one representative image is shown.

Figure 9. The first eight amino acids ofEsCBL10b are important for activation of theSOS pathway. A, Yeast (AXT3K) was trans-formed with Arabidopsis SOS1 and SOS2 incombination with AtCBL10 (A10), EsCBL10a(E10a), EsCBL10b (E10b), or the chimericgenes. Red, E10a protein; blue, E10b protein; I,insertion of seven amino acids in E10a; HD,hydrophobic domain; V, variable sequence ofamino acids; Ca, calcium-binding domains; S,Ser phosphorylation site. Serial decimal dilu-tions of yeast cells were spotted onto controlmedia or media containing 125 mM NaCl. Twoindependently transformed colonies wereassayed in three biological replicates; onerepresentative image is shown. B, Transcriptaccumulation. 18S, Loading control. One rep-resentative image of three replicates is shown.C, Amino acids in the N3 fragment werealigned and color coded based on side chainproperties: blue, nonpolar; magenta, polar;green, negatively charged; orange, positivelycharged.


Monihan et al.

Dow



Figure 10. The hydrophobic domain of EsCBL10a is important for complementation of Atsos3. Chimeric EsCBL10a (E10a)/EsCBL10b (E10b) proteins were expressed in Atsos3 and growth in the absence (Control) and presence of salt (75 mM NaCl) wasmonitored. A, Schematic representation of the chimeric proteins. Red, E10a protein; blue, E10b protein; I, insertion of sevenamino acids in E10a; HD, hydrophobic domain; V, variable sequence of amino acids; Ca, calcium-binding domains; S, Serphosphorylation site. B, Photographs of wild type (WT), Atsos3, and Atsos3 expressing E10a, E10b, or the chimeric proteins.Bar = 1 cm for all images. C, Freshweight and length of the primary root weremeasured to quantify growth. Data aremeans6 SE of



Dow



were measured and their protein activities were ex-amined. Both EsCBL10a and EsCBL10b function in theresponse of Eutrema to salt and likely have distinctroles (Fig. 3). Both genes complemented Atcbl10, dem-onstrating functional conservation between CBL10 inArabidopsis and Eutrema (Fig. 4). However, evidencefor a divergence in protein activity was seen in thedifferent abilities of EsCBL10a and EsCBL10b to activatethe SOS pathway and complement Atsos3 (Figs. 5, 7,and 13). EsCBL10a complemented Atsos3, suggestingthat it might share overlapping functions with EsSOS3(Fig. 7). However, EsCBL10a and EsSOS3 complementAtsos3 in an organ-specific manner and have differingabilities to complement Atcbl10 and activate the SOSpathway, suggesting that these genes have distinctroles in response to salt (Figs. 4, 7, and 13; SupplementalFigs. S10 and S11). Our results demonstrate that, as aresult of the duplication of CBL10, Eutrema has threecalcium sensors, EsCBL10a, EsCBL10b, and EsSOS3,with both overlapping and distinct functions involvingmultiple signaling pathways (Fig. 13).

Because the duplication in Eutrema expanded thecomplexity of calcium sensors that contribute to its salttolerance, and because Arabidopsis is more salt-sensitive and contains only one CBL10 gene, we askedif the combined action of more than one Eutrema cal-cium sensor could confer salt tolerance to Arabidopsis.Even though EsCBL10a and EsCBL10b are necessary forEutrema’s salt tolerance (Fig. 3), expression of EsC-BL10a and EsCBL10b in combination in Atcbl10was notconsistently sufficient to confer increased salt toleranceto Arabidopsis. When EsCBL10a and EsSOS3 wereexpressed together in Atsos3, root growth increasedbeyond that of the wild type and Atsos3 expressing theindividual genes, suggesting that the combined func-tions of EsCBL10a and EsSOS3 are important for rootgrowth in the presence of salt and can confer salt tol-erance to Arabidopsis (Fig. 12). Because the compo-nents of the EsCBL10 signaling pathways are likelylimiting in Arabidopsis, future studies in which thesecomponents are identified and expressed in Arabi-dopsis might be necessary for increasing salt tolerance.

Multiple Mechanisms Underlie Differences in EsCBL10Gene Function

Differences in CBL10 function can be achievedthrough changes in protein localization, affinity forcalcium, interaction with and activation of a targetprotein, or phosphorylation by a target protein. We

reasoned that the enhanced ability of EsCBL10b to ac-tivate the SOS pathway might be due to a stronger in-teraction with SOS2 and tested this hypothesis usingyeast two-hybrid assays. EsCBL10b interacted withAtSOS2 more strongly than EsCBL10a, but at a levelsimilar to AtCBL10 (Fig. 6). This result suggests thatinteraction alone cannot explain pathway activationand that additional factors must play a role.

Experiments with chimeric proteins revealed that theability of EsCBL10b to strongly activate the SOS path-way resides in the first eight amino acids of the protein(Fig. 9). The AtCBL10 hydrophobic domain is largeenough to span the membrane (Quan et al., 2007);however, this would mean that the amino acids thatunderlie EsCBL10b’s activation of the SOS pathwaywould be extracellular. If the EsCBL10b hydrophobicdomain associates with the membrane via a hydro-phobic loop (Winter et al., 2009; Hernández-Gras andBoronat, 2015; Xu et al., 2016), the amino acids 59 to thehydrophobic domain would reside in the cytosol andmight influence subcellular localization and/or inter-action with a target protein. In EsCBL10a, seven aminoacids are present in this region, suggesting that the re-duced ability of EsCBL10a to activate the SOS pathwaymight be due to this insertion (Fig. 9).

Because EsCBL10a only weakly activated the SOSpathway and showed little interaction with SOS2, yeasttwo-hybrid assays were performed to determine if in-teraction with a different kinase underlies EsCBL10a’sability to complement Atsos3. Four CIPK proteins wereidentified that interact with EsCBL10a but notEsCBL10b or AtCBL10 (Fig. 8). Three of the four alsointeract with AtSOS3, suggesting that EsCBL10a mightbe performing both SOS3-like and SOS3-independentfunctions (Fig. 8). Two of the identified CIPK proteins,AtCIPK16 and AtCIPK6, have known roles in salt tol-erance (Tsou et al., 2012; Roy et al., 2013); both havebeen shown to interact with and activate the potassiumtransporter AtAKT1 (Lee et al., 2007). In addition,AtCIPK6 has been shown to interact with AtSOS3 torecruit the potassium transporter AtAKT2 to theplasma membrane (Held et al., 2011). These resultssuggest that EsCBL10a could interact with AtCIPK6and/or AtCIPK16 proteins to regulate potassium levelsin Atsos3, thereby alleviating some of the toxic effects ofsodium. A homolog of AtCIPK16 has been identified inEutrema but the role of EsCIPK16 in plant responses tosalt has not been explored (Amarasinghe et al., 2016).Future studies focusing on EsCBL10a interaction withEsCIPK proteins and down-regulation of EsCIPK geneexpression will link the associated proteins to salt

Figure 10. (Continued.)at least 24 seedlings per genotype grown in three independent experiments. For all graphs, different letters indicate significantdifferences between genotypes (two-way ANOVA, Tukey-Kramer HSD, P# 0.05). D, Amino acids in the amino terminus of E10a,E10b, AtCBL10 (A10), and AtSOS3 (A3) proteins were aligned and color coded based on side chain properties: blue, nonpolar;magenta, polar; green, negatively charged; orange, positively charged. Major element, Portion of the N2 fragment conferring fullability to complement Atsos3; Minor element, portion of the N1 fragment conferring partial ability to complement Atsos3; blackboxes, amino acid differences that might underlie complementation.


Monihan et al.

Dow





tolerance. EsCBL10a also interacted with AtCIPK13and AtCIPK18 (Fig. 8); however, a role for these genesin salt tolerance has not been reported. Single anddouble mutant phenotypes (based on the phylogeneticrelatedness of the two genes) were assessed; however,salt-sensitive phenotypes were not detected (data notshown). These results suggest that these genes do notfunction with EsCBL10a to complement Atsos3 or thatadditional redundant genes prevent the manifestationof a salt-sensitive phenotype.EsCBL10 chimeric proteins revealed that the EsCBL10a-

specific function resides in the amino terminus within thehydrophobic domain. While little is known about thefunction of the hydrophobic domain in EsCBL10, studiesin Arabidopsis and poplar (Populus trichocarpa) suggestthat the CBL10 hydrophobic domain is important forsubcellular localization. Removal of the domain fromAtCBL10 prevented it from recruiting AtSOS2 to theplasma membrane (Quan et al., 2007), while its removalfrom the duplicated CBL10 proteins in poplar preventedtheir localization to the vacuolar membrane (Tang et al.,2014). When compared with AtCBL10 and EsCBL10b,two differences in amino acid sequence were observed inthe hydrophobic domain of EsCBL10a (major element;Fig. 10). A second element was identified in EsCBL10a

that confers partial function (minor element; Fig. 10). Thechimeric protein containing the EsCBL10b hydrophobicdomain and the EsCBL10a variable domain was able torestore rosette growth towild-type-like levels but not rootlength (bN2; Fig. 10). Because of its proximity to thehydrophobic domain, this region might influence locali-zation of the protein. Alternatively, while the three-dimensional structure of other CBL proteins indicatesthat this regionwould be unlikely to directly interact witha protein partner, it might regulate initial recognition ofthat protein. Future studies will involve localization oftagged EsCBL10 proteins driven by their native pro-moters in Eutrema. However, the low abundance of theCBL10 protein when expressed from the native promoterand its uncertain localization using a constitutive pro-moter (Kim et al., 2007; Quan et al., 2007; Tang et al., 2014)suggest that these experimentsmight prove to be difficult.

Changes in CBL10 Expression May Have Also Increasedthe Complexity of Calcium-Mediated Signaling inEutrema’s Response to Salt

Our results indicate that the expression patterns ofthe EsCBL10 genes diverged; the EsCBL10b transcript is

Figure 11. EsCBL10a and EsCBL10b in combination can confer salt tolerance. Atcbl10 plants strongly expressing EsCBL10a andEsCBL10b individually were crossed to generate heterozygous plants expressing both genes and growth in the absence (Control)or presence of salt (125 mM NaCl) was monitored. A, Photographs of wild type (WT), Atcbl10, and Atcbl10 expressing EsCBL10a(E10a), EsCBL10b (E10b), or both genes (abD). Bar = 1 cm for all images. B, Fresh weight and the length of the primary root weremeasured to quantify growth. Data are means6 SE of at least 24 seedlings per genotype grown in three independent experiments.For all graphs, different letters indicate significant differences between genotypes (two-way ANOVA, Tukey-Kramer HSD, P #

0.05). C, Transcript accumulation. ACTIN, Loading control. One representative image of three replicates is shown.



Dow



predominantly present in aerial tissues while the EsC-BL10a transcript is present throughout the plant (Fig. 2).Alternative splicing of AtCBL10 results in two majorvariants, only one of which encodes a functional protein

(Quan et al., 2007). Most of the EsCBL10a and EsCBL10bsplice variants identified likely encode nonfunctionalproteins due to premature stop codons or the absence ofdomains required for binding calcium (Supplemental

Figure 12. EsCBL10a and EsSOS3 in combination confer salt tolerance. Atsos3 plants expressing EsCBL10a and EsSOS3 indi-vidually were crossed to generate heterozygous plants expressing both genes and growth in the absence (Control) or presence ofsalt (75 mM NaCl) was monitored. A, Photographs of wild type (WT), Atsos3, and Atsos3 expressing EsCBL10a (E10a), EsSOS3(E3), or both genes (a3D). Bar = 1 cm for all images. B, Fresh weight and the length of the primary root were measured to quantifygrowth. Data are means 6 SE of at least 24 seedlings per genotype grown in three independent experiments. For all graphs,different letters indicate significant differences between genotypes (two-way ANOVA, Tukey-Kramer HSD, P # 0.05). C, Tran-script accumulation. ACTIN, Loading control. One representative image of three replicates is shown.

Figure 13. The duplication of EsCBL10 in-creased the complexity of signaling in responseto salt. In Arabidopsis, AtSOS3 and AtCBL10interact with and activate the protein kinaseAtSOS2, which, in turn, activates the sodium-proton exchanger AtSOS1 (AtSOS pathway) inthe plasma membrane (PM). AtSOS1 thentransports sodium out of the cell, preventing itstoxic accumulation. AtSOS3 and AtCBL10 haveadditional functions outside of activation of theAtSOS pathway. Homologs were identified inEutrema and one gene, EsCBL10, has been du-plicated. EsCBL10b has an enhanced ability toactivate the SOS pathway and complements theAtcbl10 salt-sensitive phenotype. EsCBL10acomplements both the Atcbl10 and Atsos3 salt-sensitive phenotypes but only weakly activatesthe SOS pathway. EsSOS3 activates the SOSpathway and complements the Atsos3 salt-sensitive phenotype.


Monihan et al.

Dow




Fig. S1). However, one EsCBL10b variant, abundant inthe lower band in the reverse transcription-PCR gel(Fig. 2), lacks the hydrophobic domain (thought totarget AtCBL10 to the plasma membrane; Quan et al.,2007) but has all four calcium-binding domains, sug-gesting that it may encode a functional protein with adifferent subcellular localization (Supplemental Fig.S1). Once functional variants have been identified, itmay be possible to determine if transcriptional regula-tion of the EsCBL10 genes further expanded calciumsignaling in Eutrema in response to salt by linkingtranscript presence and abundance to function.

Evolutionary Models May Explain Differences in EutremaCBL10 Function

Several models have been developed to describe howgenes diverge in function so that both are maintained(Flagel and Wendel, 2009; Hahn, 2009; Innan andKondrashov, 2010; Kondrashov, 2012). Of these, twomight best explain how EsCBL10a and EsCBL10b func-tion diverged. Neofunctionalization postulates that,following a gene duplication, one gene maintains theoriginal function, while the other gene, free from se-lective constraints, develops a novel, advantageousfunction (Ohno, 1970). In the case of the EsCBL10 genes,EsCBL10b might have maintained the original function(activation of the SOS pathway) while EsCBL10a de-veloped a novel function (complementation of Atsos3).Subfunctionalization (duplication-degeneration-com-plementation) postulates that the original gene (beforethe duplication) had multiple functions that were laterpartitioned between the duplicated genes (Force et al.,1999). In the case of the EsCBL10 genes, EsCBL10 (beforethe duplication) would have had both functions (acti-vation of the SOS pathway and complementation ofAtsos3). Following the duplication, one function wasmaintained in EsCBL10b (activation of the SOS path-way) and one function (complementation of Atsos3)was lost. Similarly, EsCBL10a maintained one function(complementation of Atsos3) and lost the other function(activation of the SOS pathway). Once the functionswere partitioned, each gene could acquire changes thatmight enhance its function; something that might nothave been possible when one gene performed bothfunctions (escape from adaptive conflict; Hughes, 1994;DesMarais and Rausher, 2008). To distinguish betweenthese models and determine if EsCBL10a developed thefunction that complements Atsos3 before the duplica-tion (subfunctionalization) or after the duplication(neofunctionalization), functional data for CBL10 fromadditional species will be required.

MATERIALS AND METHODS

Plant Material

Arabidopsis (Arabidopsis thaliana Columbia-0) and Eutrema (Eutrema salsu-gineum Shangdong) were used as the wild types for this study. An Arabidopsis

CBL10 T-DNA insertion line (SALK_056042) was obtained from the Arabi-dopsis Biological Resource Center and a homozygous mutant was generated(Monihan et al., 2016). The Arabidopsis sos1-1 (Wu et al., 1996), sos2-2 (Zhuet al., 1998), and sos3-1 (Liu and Zhu, 1997) ethyl methanesulfonate mutantswere obtained from Dr. Jian-Kang Zhu (Purdue University).

Plant Growth and Salt Assays

To monitor seedling shoot and root development, Arabidopsis seeds weresown on 0.53 Murashige and Skoog (MS) medium (PhytoTechnology Labo-ratories) containing 2.5 mM MES, 2% (w/v) Suc, and 1% (w/v) agar (A8678;Sigma). Eutrema seeds were sown on Eutrema BasalMedium (EBM) containing3 mM calcium nitrate, 1.5 mM magnesium sulfate, 1.25 mM potassium phos-phate, 13 micronutrient solution (100 mM boric acid, 0.1 mM cobalt chloride,0.1 mM cupric sulfate, 100 mM ferrous sulfate, 100 mM manganese sulfate, 5 mM

potassium iodine, 1 mM sodium molybdate, 30 mM zinc sulfate, and 100 mM

disodium EDTA), 2.5 mM MES, 2% (w/v) Suc, and 1% (w/v) agar. All mediawas brought to a final pH of 5.7 (adjusted with potassium hydroxide). Plateswith seed were incubated at 4°C in the dark for 2 d (Arabidopsis) or 4 d(Eutrema) to break dormancy and transferred to a growth chamber at 21°Cunder a 16-h-light/8-h-dark photoperiod with light provided by PhillipsF32T8/TL841 bulbs (135 mmol m22 s21).

For salt assays, seeds were germinated on media without NaCl for 4 d(Arabidopsis) or 5 d (Eutrema) after which seedlings were transferred to mediawithout or with the indicated concentration of NaCl. NaCl concentrations werechosen formaximal differences in growth between thewild type andmutants oramiRNA lines. Values for fresh weight represent the entire seedling weight;values for root length represent the amount of growth from the time of seedlingtransfer to the end of the assay, 10 d for Arabidopsis and 14 d for Eutrema.

To monitor vegetative (rosette) and reproductive development, Arabidopsisand Eutrema were grown in SunGro Sunshine LC1 soil mix (SunGrow Horti-culture) and watered every 3 d with 0.53 Hoagland solution (Hoagland andArnon, 1938) with cobalt chloride in place of cobalt nitrate and a final pH of 5.7(adjusted with potassium hydroxide). To monitor vegetative development, Ara-bidopsis and Eutrema seeds were stratified for 4 d at 4°C in the dark to breakdormancy. To monitor reproductive development, seeds from both species werestratified for 3weeks at 4°C in the dark to break dormancy andpromotefloweringin Eutremawithout vernalization.After cold treatment, plantswere transferred toa growth chamber at 21°C under a 16-h-light/8-h-dark photoperiod. For saltassays, NaCl was added to 0.53Hoagland solution and plants were treated withNaCl in 50 mM increments every 3 d until the indicated, final concentration wasreached. To monitor the effect of salt, NaCl treatments were started 1 week aftergermination and continued for 3 weeks (vegetative development) or started3 weeks after germination, at the start of inflorescence development, and con-tinued for 4 weeks (reproductive development).

Transcript Analysis

EsCBL10a and EsCBL10b were identified from the Department of EnergyJoint Genome Institute genome assembly, JGI unmasked v1. EsCBL10a(Thhalv10026019m, chromosome 7) was identified on scaffold 1, 2936177-2934528, and EsCBL10b (Thhalv10028908m, chromosome 6) was identified onscaffold 3, 7161515-7163137.

Arabidopsis andEutrema leaves and roots of 2-week-old seedlings grownonMS or EBM plates and open flowers (stage 14; Smyth et al., 1990) from plantsgrown in soil were collected for transcript analysis. To determine transcriptlevels of transgenes in Arabidopsis, 2-week-old seedlings grown on MS plateswere collected. RNA was isolated using the Qiagen RNeasy Plant Mini Kit,treated with TURBO DNase (Invitrogen), purified (RNeasy MinElute CleanupKit; Qiagen), and used to synthesize cDNA (M-MLV Reverse Transcriptase;Promega). All PCR reactions were performed using Recombinant Taq Poly-merase (Invitrogen) and were analyzed and compared in the linear range of theamplification (e.g. 22 cycles forActin and 30 cycles for CBL10) using the primersprovided in Supplemental Table S1. Multiple bands (two to three) were iden-tified for the endogenous CBL10 genes. Products from each band were clonedinto pGEM-T Easy (Promega) and 10 clones were sequenced per band toidentify alternatively spliced variants.

amiRNAs

To reduce expression of EsCBL10a and EsCBL10b individually and in com-bination, amiRNAs were designed using the Web MicroRNA Designer site



Dow







(http://wmd3.weigelworld.org; Schwab et al., 2006; Ossowski et al., 2008).Two 21-bp sequences per target (EsCBL10a, EsCBL10b, and EsCBL10a/EsCBL10b) were identified. Primers (Supplemental Table S2) containing theamiRNA sequences were generated and the amiRNA sequence was incorpo-rated into the AtMIR319a precursor gene located on the pRS300 vector (Dr.DetlefWeigel, Max Planck Institute for Developmental Biology; Ossowski et al.,2008). The AtMIR319a gene containing the desired amiRNA sequence wasdigested with EcoRI and XbaI and subcloned into the corresponding site of theplant binary vector pEZT-NL (Drs. Sean Cutler and David W. Ehrhardt, Car-negie Institution of Washington) downstream of the CaMV 35S promoter.Agrobacterium tumefaciens EH105 containing the binary vector was used totransform Eutrema via the floral dip method (Clough and Bent, 1998). T1 seedwas germinated on soil for 1.5 weeks and then sprayed three times with Basta(100 mg/L glufosinate ammonium; Rely 200 Herbicide; Bayer Crop Science) at3-d intervals. T1 lines with antibiotic resistance were subsequently transferredto pots and grown to collect T2 seed. Single insertion lines were identified byscreening T2 seed on EBM plates containing 7.5 mg L21 glufosinate ammonium(Santa Cruz Biotechnology) and selecting lines with 75% resistance. Homozy-gous lines were identified by screening T3 seed on EBM plates with glufosinateammonium and selecting lines with 100% resistance. Salt assays were per-formed as described above.

The efficiency of amiRNA targeting was determined using reverse tran-scription quantitative PCR. RNAwas isolated from 3-week-old seedlings usingTrizol (Invitrogen). RNA was treated with RQ1 DNase (Promega), purifiedusing the RNeasy Plant Mini Kit (Qiagen), and used to synthesize cDNA(M-MLV Reverse Transcriptase; Promega). Reverse transcription quantitativePCR was performed using the LightCycler Fast Start DNA MasterPLUS SYBRGreen I Kit (Roche) in the LightCycler 1.5 instrument (Roche). Melting-curveanalysis was performed to verify that a single PCR product was amplified. CT

values were calculated using the LightCycler software 4.0 package (Roche). Therelative fold change in expression was determined using the calculation 22DDCT

(Schmittgen and Livak, 2008). All primers are provided in SupplementalTable S1.

Complementation Assays

The functions of EsCBL10a, EsCBL10b, and EsSOS3 were analyzed byexpressing each gene in the Atcbl10 and Atsos3 mutants. Full-length codingsequences without stop codons were amplified from cDNA and cloned intopGEM-T Easy (Promega). All PCR amplifications were performed using Phu-sion High-Fidelity DNA Polymerase (ThermoFisher Scientific); primers areprovided in Supplemental Table S1. AtCBL10, EsCBL10a, and EsCBL10b weredigested with XhoI and BamHI and EsSOS3 was digested with XhoI and EcoRI.All genes were subcloned into the corresponding site of the plant binary vectorpEZT-NL containing the CaMV 35S promoter. A. tumefaciens LBA4404 con-taining the binary vector was used to transformAtcbl10 andAtsos3 via the floraldip method (Clough and Bent, 1998). A similar strategy was used to generatesingle insertion, homozygous lines as described above for the amiRNA lines,exceptMSmediumwas used in place of EBM.AtSOS3 expressed inAtcbl10 andAtsos3 downstream of the CaMV 35S promoter were provided by Dr. Yan Guo(China Agricultural University). To determine if coexpression of more than oneEutrema calcium sensor confers salt tolerance to Arabidopsis, plants expressingEsCBL10a, EsCBL10b, and EsSOS3 individually were crossed to generate het-erozygous plants expressing two genes. Salt assays were performed asdescribed above.

Yeast Salt Assays

To monitor the ability of EsCBL10a and EsCBL10b to activate the SOSpathway, yeast (Saccharomyces cerevisiae) strain AXT3K (ena1::HIS3::ena4, nha1::LEU2, and nhx1::KanMX4; Quintero et al., 2002) containing the pYPGE15plasmid with AtSOS1 or EsSOS1 (Jarvis et al., 2014) and the pFL32T plasmidwith AtSOS2 and AtSOS3 or the pFLE3E2T plasmid with EsSOS2 and EsSOS3(Jarvis et al., 2014)weremodified to express theCBL10 genes. In order to replaceAtSOS3 and EsSOS3with the CBL10 genes from Arabidopsis and Eutrema, thefull-length coding sequences of AtCBL10, EsCBL10a, and EsCBL10b were am-plified from cDNA (Phusion High-Fidelity DNA polymerase; ThermoFisherScientific; primers are provided in Supplemental Table S1), cloned into pGEM-TEasy (Promega), digested with XhoI and NotI, and subcloned into the corre-sponding site of the pDR195 vector (Dr. Alonso Rodriguez-Navarro; Rentschet al., 1995). The plasmid was digested with AgeI and NotI and a fragmentcontaining the CBL10 gene was subcloned into the corresponding site of the

pFL32T plasmid in place of AtSOS3 to be expressed with AtSOS2 or thepFLE3E2T plasmid in place of EsSOS3 to be expressed with EsSOS2. Plasmidscontaining AtSOS2 or EsSOS2 but lacking the CBL10 and SOS3 genes weregenerated as described previously (Jarvis et al., 2014). TransformedAXT3K cellswere selected on synthetic dropout medium lacking uracil and Trp (Clontech/TaKaRa) containing yeast nitrogen base without amino acids (VWR). Saltassays were carried out in alkali cation-free medium (AP; Rodríguez-Navarroand Ramos, 1984) containing 1 mM KCl with the designated concentrations ofNaCl and cultured at 30°C for 4 d.

To monitor the level of CBL10 expression in AXT3K, spheroplasts weregenerated by removing the cell wall with Lyticase (Sigma; Yılmaz et al., 2012).RNA was extracted from spheroplasts using the TRI REAGENT (Invitrogen;Chomczynski and Sacchi, 1987, 2006; Yılmaz et al., 2012), treated with TURBODNase (ThermoFisher Scientific), and purified using the RNeasy MiniEluteCleanup Kit (Qiagen). cDNA was synthesized using SuperScript III ReverseTranscriptase (Life Technologies) and RNA was removed using RNaseH (LifeTechnologies). All PCR reactions were performed using Recombinant TaqPolymerase (Invitrogen); primers are provided in Supplemental Table S1.

Yeast Two-Hybrid Screens

EsCBL10a-interacting kinases were identified using yeast two-hybridscreens. Each gene to be tested for yeast two-hybrid interaction was ampli-fied by PCR using Phusion High-Fidelity DNA Polymerase (ThermoFisherScientific); primers are provided in Supplemental Table S1. For CIPK genes 1, 2,4, 6, 7, 9 to 11, 13, 18, 19, 21, and 24 (SOS2), the PCR products were digestedwithEcoRI and BamHI and cloned into the corresponding site of the pGADT7 andpGBKT7 vectors (Clontech/TaKaRa), which allow for expression of the genesas fusion proteins with the GAL4 DNA-activation domain or the GAL4 DNA-binding domain, respectively. For CIPK genes 3, 5, 8, 12, 14 to 17, 20, 22, 23, and26, the PCR products were digested with XmaI and XhoI and cloned into thecorresponding site of pGADT7. pGBKT7 does not have a XhoI site but the SalIsite produces the same overhang, so each CIPK gene digested with XmaI andXhoI was cloned into pGBKT7 digested with XmaI and SalI. For CIPK25, whichcontains an internalXhoI site, the PCR product was digested withXmaI and SalIand cloned into the corresponding site of pGBKT7 and into pGADT7 digestedwith XmaI and XhoI. The pGADT7 clones were transformed into S. cerevisiaestrain Y2HGold (Clontech/TaKaRa) except for the AtSOS3 clone, which hadbeen previously transformed into S. cerevisiae strain AH109. The pGBKT7clones were transformed into S. cerevisiae strain Y187 (Clontech/TaKaRa). Yeastwere mated and grown on SDmediumminus Leu and Trp (Clontech/TaKaRa)to select for diploid yeast expressing both constructs. To determine interaction,serial decimal dilutions of yeast colonies were grown on SDmediumminus Leuand Trp and without His and adenine or with the addition of 0.5 mM 3-amino-1,2,4-triazole (Sigma) and incubated for 5 d.

Chimeric Proteins

To generate chimeric EsCBL10 for expression in yeast, EsCBL10a andEsCBL10b were cloned into pDR195 (cloning described in “Yeast Salt Assays”above). For the first set of chimeric proteins, HindIII was used to generate twosites in the plasmid, one in the vector and the other within each gene. Theresulting fragments containing the amino terminus of each gene were thencloned into the vector containing the carboxy terminus of the other gene. Theplasmids were digested with AgeI and NotI and a fragment containing thechimeric gene was subcloned into the corresponding site of the pFL32T plasmidto be expressed withAtSOS2 in AXT3K. For the second set of chimeric proteins,each gene fragment (aN2, aC2, bN2, and bC2) was PCR amplified using Phu-sion High-Fidelity DNA Polymerase (ThermoFisher Scientific); primers areprovided in Supplemental Table S1. The primers were designed so that the aN2and bC2 and the bN2 and aC2 fragments would overlap. The fragments wereamplified by PCR and the full-length chimeric gene cloned into pGEM-T Easy(Promega). The plasmid was then digested with XhoI and NotI and a fragmentcontaining the chimeric gene was subcloned into the corresponding site of thepDR195 vector. The resulting plasmid was digested with AgeI andNotI and thefragment containing the chimeric gene was subcloned into the correspondingsites of the pFL32T plasmid to be expressed with AtSOS2 in AXT3K. For thethird set of chimeric proteins, EsCBL10a and EsCBL10b cloned into pDR195were transformed into dam2/dcm2 competent cells (New England Biolabs) toget a methylation-free plasmid. The resulting plasmids were digested withAgeIand XbaI and the fragments containing the amino terminus of one gene werecloned into the plasmid containing the carboxy terminus of the other gene. The


Monihan et al.

Dow



http://wmd3.weigelworld.org









resulting plasmid was digested withAgeI andNotI and the fragment containingthe chimeric gene was subcloned into the corresponding site of the pFL32Tplasmid to be expressed with AtSOS2 in AXT3K.

For expression in Arabidopsis, each chimeric gene was amplified from theyeast plasmids and cloned into pGEM-T Easy (Promega). All PCR amplifica-tions were performed using Phusion High-Fidelity DNA Polymerase (Ther-moFisher Scientific); primers are provided in Supplemental Table S1. Eachchimeric gene was digested with XhoI and BamHI and subcloned into the cor-responding site of the plant binary vector pEZT-NL containing the CaMV 35Spromoter. Transformation of Arabidopsis and generation of homozygous lineswere performed as described above in “Complementation Assays.”

Statistical Analysis

To determine significant differences in growth, experiments were organizedand analyzed as a randomized complete block design with genotypes and saltconcentrations as treatments and individual experiments as replicates. Treat-ment effects were assessed using a full-factorial mixed-model ANOVA in JMP,Version 11 (SAS Institute; 1989–2007). In these analyses, treatments were con-sidered fixed effects and replicates random effects. The normality of the dis-tributions of all dependent variables was analyzed by examining a plot of theresiduals from a full-factorial ANOVA of untransformed data. A Shapiro-Wilktest (Shapiro and Wilk, 1965) was performed to assess normality, and Bartlett’s(Bartlett, 1937) and Levene’s (Levene, 1960) tests were performed to evaluatethe homogeneity of variance. Based on the pattern of distribution and the re-sults of these tests, a nonparametric approach was used to analyze the data.Data were rank transformed using Microsoft Excel (function: RANK) followedby an ANOVA and Tukey’s HSD test for multiple comparisons of means(Conover and Iman, 1981). The HSD values from rank-based ANOVA werethen applied to the actual means for each measurement (i.e. not the ranks usedin ANOVA). Statistical significance was assigned at P # 0.05, and all tests ofsignificance were two sided.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL datalibraries under accession numbers At4g33000.2 (AtCBL10), At5g24270(AtSOS3), ESQ53857 (EsCBL10a, Thhalv10026019m), ESQ38534 (EsCBL10b,Thhalv10028908m), and ESQ31966 (EsSOS3, Thhalv10004891m).

Supplemental Data

The following supplemental materials are available.

Supplemental Figure S1. AtCBL10, EsCBL10a, and EsCBL10b are alterna-tively spliced.

Supplemental Figure S2. AmiRNA annealing sites and specificity

Supplemental Figure S3. Reduced expression of EsCBL10a and EsCBL10bresults in decreased root growth.

Supplemental Figure S4. EsCBL10a and EsCBL10b complement the Atcbl10fertilization defect.

Supplemental Figure S5. EsCBL10b interaction with AtSOS2 is orientationspecific.

Supplemental Figure S6. Orientation of the yeast two-hybrid affectsCBL10 interaction with the CIPK proteins.

Supplemental Figure S7.. Chimeric EsCBL10 proteins.

Supplemental Figure S8. The chimeric EsCBL10 genes are expressed in theAtcbl10 and Atsos3 mutants.

Supplemental Figure S9. All chimeric proteins complement the Atcbl10salt-sensitive phenotype.

Supplemental Figure S10. EsSOS3 activates the Arabidopsis and EutremaSOS pathways.

Supplemental Figure S11. EsSOS3 does not complement the Atcbl10 salt-sensitive phenotype.

Supplemental Table S1. Primers.

Supplemental Table S2. Primers used for generating the amiRNAconstructs.

ACKNOWLEDGMENTS

We thank Brian M. Ortega and Katerina A. Oskolkoff (School of PlantSciences, University of Arizona) for technical assistance, Dr. Yan Guo (State KeyLaboratory of Plant Physiology and Biochemistry, China Agricultural Univer-sity) for providing Atcbl10 and Atsos3 lines expressing AtSOS3, and Dr. DetlefWeigel (Department of Molecular Biology, Max Planck Institute for Develop-mental Biology) for providing the amiRNA vector. We thank Drs. Ramin Yade-gari and Steven E. Smith (School of Plant Sciences, University of Arizona) forconsultations on experiments and statistical analyses.

Received November 8, 2018; accepted December 16, 2018; published January 3,2019.

LITERATURE CITED

Akaboshi M, Hashimoto H, Ishida H, Saijo S, Koizumi N, Sato M, ShimizuT (2008) The crystal structure of plant-specific calcium-binding proteinAtCBL2 in complex with the regulatory domain of AtCIPK14. J Mol Biol377: 246–257

Amarasinghe S, Watson-Haigh NS, Gilliham M, Roy S, Baumann U (2016)The evolutionary origin of CIPK16: A gene involved in enhanced salttolerance. Mol Phylogenet Evol 100: 135–147

Amtmann A (2009) Learning from evolution: Thellungiella generates newknowledge on essential and critical components of abiotic stress tolerancein plants. Mol Plant 2: 3–12

Bartlett MS (1937) Properties of sufficiency and statistical tests. Proc R SocLond A Math Phys Sci 160: 268–282

Bressan RA, Zhang C, Zhang H, Hasegawa PM, Bohnert HJ, Zhu JK (2001)Learning from the Arabidopsis experience: The next gene search para-digm. Plant Physiol 127: 1354–1360

Chomczynski P, Sacchi N (1987) Single-step method of RNA isolation byacid guanidinium thiocyanate-phenol-chloroform extraction. Anal Bio-chem 162: 156–159

Chomczynski P, Sacchi N (2006) The single-step method of RNA isolationby acid guanidinium thiocyanate-phenol-chloroform extraction: Twenty-something years on. Nat Protoc 1: 581–585

Clough SJ, Bent AF (1998) Floral dip: A simplified method forAgrobacterium-mediated transformation of Arabidopsis thaliana. Plant J16: 735–743

Conover WJ, Iman RL (1981) Rank transformations as a bridge betweenparametric and nonparametric statistics. Am Stat 35: 124–129

Des Marais DL, Rausher MD (2008) Escape from adaptive conflict afterduplication in an anthocyanin pathway gene. Nature 454: 762–765

FAO (2002) The salt of the earth: hazardous for food production. http://www.fao.org/worldfoodsummit/english/newsroom/focus/focus1.htm

Flagel LE, Wendel JF (2009) Gene duplication and evolutionary novelty inplants. New Phytol 183: 557–564

Force A, Lynch M, Pickett FB, Amores A, Yan YL, Postlethwait J (1999)Preservation of duplicate genes by complementary, degenerative muta-tions. Genetics 151: 1531–1545

Gonzales-Vigil E, Bianchetti CM, Phillips GN Jr, Howe GA (2011) Adap-tive evolution of threonine deaminase in plant defense against insectherbivores. Proc Natl Acad Sci USA 108: 5897–5902

Guo Y, Halfter U, Ishitani M, Zhu JK (2001) Molecular characterization offunctional domains in the protein kinase SOS2 that is required for plantsalt tolerance. Plant Cell 13: 1383–1400

Hahn MW (2009) Distinguishing among evolutionary models for the main-tenance of gene duplicates. J Hered 100: 605–617

Halfter U, Ishitani M, Zhu JK (2000) The Arabidopsis SOS2 protein kinasephysically interacts with and is activated by the calcium-binding proteinSOS3. Proc Natl Acad Sci USA 97: 3735–3740

Hanikenne M, Talke IN, Haydon MJ, Lanz C, Nolte A, Motte P, KroymannJ, Weigel D, Krämer U (2008) Evolution of metal hyperaccumulationrequired cis-regulatory changes and triplication of HMA4. Nature 453:391–395

Held K, Pascaud F, Eckert C, Gajdanowicz P, Hashimoto K, Corratgé-Faillie C, Offenborn JN, Lacombe B, Dreyer I, Thibaud JB, et al



Dow

















(2011) Calcium-dependent modulation and plasma membrane targetingof the AKT2 potassium channel by the CBL4/CIPK6 calcium sensor/protein kinase complex. Cell Res 21: 1116–1130

Hernández-Gras F, Boronat A (2015) A hydrophobic proline-rich motif isinvolved in the intracellular targeting of temperature-induced lipocalin.Plant Mol Biol 88: 301–311

Hoagland DR, Arnon DI (1938) The water-culture method for growingplants without soil. Calif Agric Exp Station Circular 347: 1–39

Hughes AL (1994) The evolution of functionally novel proteins after geneduplication. Proc Biol Sci 256: 119–124

Ilyas M, Hörger AC, Bozkurt TO, van den Burg HA, Kaschani F, Kaiser M,Belhaj K, Smoker M, Joosten MH, Kamoun S, et al (2015) Functionaldivergence of two secreted immune proteases of tomato. Curr Biol 25:2300–2306

Inan G, Zhang Q, Li P, Wang Z, Cao Z, Zhang H, Zhang C, Quist TM,Goodwin SM, Zhu J, et al (2004) Salt cress: A halophyte and cryophyteArabidopsis relative model system and its applicability to molecular ge-netic analyses of growth and development of extremophiles. Plant Physiol135: 1718–1737

Innan H, Kondrashov F (2010) The evolution of gene duplications: Classi-fying and distinguishing between models. Nat Rev Genet 11: 97–108

Jarvis DE, Ryu CH, Beilstein MA, Schumaker KS (2014) Distinct roles forSOS1 in the convergent evolution of salt tolerance in Eutrema salsugi-neum and Schrenkiella parvula. Mol Biol Evol 31: 2094–2107

Kim BG, Waadt R, Cheong YH, Pandey GK, Dominguez-Solis JR, SchültkeS, Lee SC, Kudla J, Luan S (2007) The calcium sensor CBL10 mediatessalt tolerance by regulating ion homeostasis in Arabidopsis. Plant J 52:473–484

Kondrashov FA (2012) Gene duplication as a mechanism of genomic adap-tation to a changing environment. Proc R Soc B Biol Sci 279: 5048–5057

Lee SC, Lan WZ, Kim BG, Li L, Cheong YH, Pandey GK, Lu G, BuchananBB, Luan S (2007) A protein phosphorylation/dephosphorylation net-work regulates a plant potassium channel. Proc Natl Acad Sci USA 104:15959–15964

Levene H (1960) Robust tests for equality of variances. In I Olkin, ed, Con-tributions to Probability and Statistics. Stanford University Press, Stan-ford, CA, pp 278–292

Lin H, Yang Y, Quan R, Mendoza I, Wu Y, Du W, Zhao S, Schumaker KS,Pardo JM, Guo Y (2009) Phosphorylation of SOS3-LIKE CALCIUMBINDING PROTEIN8 by SOS2 protein kinase stabilizes their proteincomplex and regulates salt tolerance in Arabidopsis. Plant Cell 21:1607–1619

Liu J, Zhu JK (1997) An Arabidopsis mutant that requires increased calciumfor potassium nutrition and salt tolerance. Proc Natl Acad Sci USA 94:14960–14964

Liu J, Zhu JK (1998) A calcium sensor homolog required for plant salt tol-erance. Science 280: 1943–1945

Liu J, Ishitani M, Halfter U, Kim CS, Zhu JK (2000) The Arabidopsis thalianaSOS2 gene encodes a protein kinase that is required for salt tolerance.Proc Natl Acad Sci USA 97: 3730–3734

Monihan SM, Magness CA, Yadegari R, Smith SE, Schumaker KS (2016)Arabidopsis CALCINEURIN B-LIKE10 functions independently of theSOS pathway during reproductive development in saline conditions.Plant Physiol 171: 369–379

Munns R, Tester M (2008) Mechanisms of salinity tolerance. Annu Rev PlantBiol 59: 651–681

Oh DH, Leidi E, Zhang Q, Hwang SM, Li Y, Quintero FJ, Jiang X, D’UrzoMP, Lee SY, Zhao Y, et al (2009) Loss of halophytism by interference withSOS1 expression. Plant Physiol 151: 210–222

Ohno S (1970) Evolution by Gene Duplication. Springer, BerlinOssowski S, Schwab R Weigel D (2008) Gene silencing in plants using ar-

tificial microRNAs and other small RNAs. Plant J 53: 674–690Qadir M, Quillerou E, Nangia V, Murtaza G, Singh M, Thomas RJ,

Drechsel P, Noble AD (2014) Economics of salt-induced land degrada-tion and restoration. Nat Resour Forum 38: 282–295

Qiu QS, Guo Y, Dietrich MA, Schumaker KS, Zhu JK (2002) Regulation ofSOS1, a plasma membrane Na+/H+ exchanger in Arabidopsis thaliana, bySOS2 and SOS3. Proc Natl Acad Sci USA 99: 8436–8441

Qiu QS, Barkla BJ, Vera-Estrella R, Zhu JK, Schumaker KS (2003) Na+/H+

exchange activity in the plasma membrane of Arabidopsis. Plant Physiol132: 1041–1052

Quan R, Lin H, Mendoza I, Zhang Y, Cao W, Yang Y, Shang M, Chen S,Pardo JM, Guo Y (2007) SCABP8/CBL10, a putative calcium sensor, in-teracts with the protein kinase SOS2 to protect Arabidopsis shoots fromsalt stress. Plant Cell 19: 1415–1431

Quintero FJ, Ohta M, Shi H, Zhu JK, Pardo JM (2002) Reconstitution inyeast of the Arabidopsis SOS signaling pathway for Na+ homeostasis.Proc Natl Acad Sci USA 99: 9061–9066

Rentsch D, Laloi M, Rouhara I, Schmelzer E, Delrot S, Frommer WB (1995)NTR1 encodes a high affinity oligopeptide transporter in Arabidopsis.FEBS Lett 370: 264–268

Rodríguez-Navarro A, Ramos J (1984) Dual system for potassium transportin Saccharomyces cerevisiae. J Bacteriol 159: 940–945

Roy SJ, Huang W, Wang XJ, Evrard A, Schmöckel SM, Zafar ZU, Tester M(2013) A novel protein kinase involved in Na+ exclusion revealed frompositional cloning. Plant Cell Environ 36: 553–568

Sánchez-Barrena MJ, Martínez-Ripoll M, Zhu JK, Albert A (2005) Thestructure of the Arabidopsis thaliana SOS3: Molecular mechanism of sens-ing calcium for salt stress response. J Mol Biol 345: 1253–1264

Schmittgen TD, Livak KJ (2008) Analyzing real-time PCR data by thecomparative C(T) method. Nat Protoc 3: 1101–1108

Schwab R, Ossowski S, Riester M, Warthmann N, Weigel D (2006) Highlyspecific gene silencing by artificial microRNAs in Arabidopsis. Plant Cell18: 1121–1133

Shapiro SS, Wilk MB (1965) An analysis of variance test for normality(complete samples). Biometrika 52: 591

Shi H, Ishitani M, Kim C, Zhu JK (2000) The Arabidopsis thaliana salt tol-erance gene SOS1 encodes a putative Na+/H+ antiporter. Proc Natl AcadSci USA 97: 6896–6901

Simon S, Rühl M, de Montaigu A, Wötzel S, Coupland G (2015) Evolutionof CONSTANS regulation and function after gene duplication produced aphotoperiodic flowering switch in the Brassicaceae. Mol Biol Evol 32:2284–2301

Smyth DR, Bowman JL, Meyerowitz EM (1990) Early flower developmentin Arabidopsis. Plant Cell 2: 755–767

Tang RJ, Yang Y, Yang L, Liu H, Wang CT, Yu MM, Gao XS, Zhang HX(2014) Poplar calcineurin B-like proteins PtCBL10A and PtCBL10B regu-late shoot salt tolerance through interaction with PtSOS2 in the vacuolarmembrane. Plant Cell Environ 37: 573–588

Tsou PL, Lee SY, Allen NS, Winter-Sederoff H, Robertson D (2012) An ER-targeted calcium-binding peptide confers salt and drought tolerancemediated by CIPK6 in Arabidopsis. Planta 235: 539–552

Winter U, Chen X, Fasshauer D (2009) A conserved membrane attachmentsite in alpha-SNAP facilitates N-ethylmaleimide-sensitive factor (NSF)-driven SNARE complex disassembly. J Biol Chem 284: 31817–31826

Wu SJ, Ding L, Zhu JK (1996) SOS1, a genetic locus essential for salt toler-ance and potassium acquisition. Plant Cell 8: 617–627

Xu X, Song H, Qi J, Liu Y, Wang H, Su C, Shi Y, Gao GF (2016) Contributionof intertwined loop to membrane association revealed by Zika virus full-length NS1 structure. EMBO J 35: 2170–2178

Yang R, Jarvis DE, Chen H, Beilstein MA, Grimwood J, Jenkins J, Shu S,Prochnik S, Xin M, Ma C, et al (2013) The reference genome of the hal-ophytic plant Eutrema salsugineum. Front Plant Sci 4: 46

Yılmaz R, Akça O, Baloğlu MC, Öz MT, Öktem HA, Yücel M (2012) Op-timization of yeast (Saccharomyces cerevisiae) RNA isolation method forreal-time quantitative PCR and microarray analysis. Afr J Biotechnol 11:1046–1053

Zhu JK, Liu J, Xiong L (1998) Genetic analysis of salt tolerance in Arabi-dopsis: Evidence for a critical role of potassium nutrition. Plant Cell 10:1181–1191


Monihan et al.

Dow



Documents

Linking Duplication of aCalcium Sensor to Salt Tolerance...The SALT-OVERLY-SENSITIVE (SOS) pathway in Arabidopsis (Arabidopsis thaliana) functions to prevent the toxic accumulation