Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Sequence Evolution and Expression Regulation of Stress-Responsive Genes in Natural Populations of Wild Tomato

  • Iris Fischer ,

    irisfischer402@gmail.com

    Current address: INRA, UMR, 1334 AGAP, Montpellier, France

    Affiliation Section of Evolutionary Biology, Department of Biology II, University of Munich, Planegg-Martinsried, Germany

  • Kim A. Steige,

    Current address: Evolutionary Biology Centre, Uppsala University, Uppsala, Sweden

    Affiliation Section of Evolutionary Biology, Department of Biology II, University of Munich, Planegg-Martinsried, Germany

  • Wolfgang Stephan,

    Affiliation Section of Evolutionary Biology, Department of Biology II, University of Munich, Planegg-Martinsried, Germany

  • Mamadou Mboup

    Current address: Department of Plant Sciences, University of Oxford, Oxford, United Kingdom

    Affiliation Section of Evolutionary Biology, Department of Biology II, University of Munich, Planegg-Martinsried, Germany

Sequence Evolution and Expression Regulation of Stress-Responsive Genes in Natural Populations of Wild Tomato

  • Iris Fischer, 
  • Kim A. Steige, 
  • Wolfgang Stephan, 
  • Mamadou Mboup
PLOS
x

Abstract

The wild tomato species Solanum chilense and S. peruvianum are a valuable non-model system for studying plant adaptation since they grow in diverse environments facing many abiotic constraints. Here we investigate the sequence evolution of regulatory regions of drought and cold responsive genes and their expression regulation. The coding regions of these genes were previously shown to exhibit signatures of positive selection. Expression profiles and sequence evolution of regulatory regions of members of the Asr (ABA/water stress/ripening induced) gene family and the dehydrin gene pLC30-15 were analyzed in wild tomato populations from contrasting environments. For S. chilense, we found that Asr4 and pLC30-15 appear to respond much faster to drought conditions in accessions from very dry environments than accessions from more mesic locations. Sequence analysis suggests that the promoter of Asr2 and the downstream region of pLC30-15 are under positive selection in some local populations of S. chilense. By investigating gene expression differences at the population level we provide further support of our previous conclusions that Asr2, Asr4, and pLC30-15 are promising candidates for functional studies of adaptation. Our analysis also demonstrates the power of the candidate gene approach in evolutionary biology research and highlights the importance of wild Solanum species as a genetic resource for their cultivated relatives.

Introduction

Numerous efforts have been made in the last decades to understand local adaptation. This phenomenon is defined as the movement of a population towards a phenotype that leads to the highest fitness in a particular environment [1]. As protein divergence alone often cannot explain the phenotypic differences observed between populations/species, gene expression regulation has been suggested to play a key role for many cases of adaptation [2-4]. Modulation of gene expression is crucial for the survival of organisms as environmental changes require fast and specific responses. Experimental evolution studies in microorganisms revealed fast expression divergence between strains of Saccharomyces cerevisiae (yeast) [5] and Escherichia coli [6] grown in glucose-limited media. In plants, regulatory changes between domesticated crop species and their wild relatives as well as their role in adaptation have been described in Zea mays (maize) [7,8] and Oryza sativa (rice) [9]. Therefore, one way to investigate local adaptation is to study the expression and regulation of genes that provide a higher fitness under stress conditions. Whole transcriptome analysis has largely been done using microarrays for investigating expression differences in natural populations of model species such as Drosophila melanogaster [10,11], to study host shifts in D. mojavensis [12] or to analyze invasive plant species such as Ambrosia artemisiifolia (common ragweed) [13] and Cirsium arvense (Canada thistle) [14]. Recently, sequencing of whole transcriptomes/exomes additionally allowed for large-scale gene expression analysis in different populations or cultivars, e.g. in D. melanogaster [15], D. mojavensis [16], or Citrullus lanatus (watermelon) [17]. However, expression differences have also been analyzed in more detail for particular candidate genes, e.g. cold responsive genes in wild tomato (Solanum sp.) [18] or genes involved in root architecture in monkey-flower (Mimulus guttatus) [19]. Another question that remains to be answered is to which degree regulatory divergence is adaptive [20]. Most analyses are still focused on the evolution of coding sequences, but examples attributing adaptation to regulatory changes have increased over the last few years [2,21,22].

Terrestrial plants are usually sessile during their life cycle and drought and cold stress are the major abiotic constraints they are facing. Both types of stress have adverse effects on plant growth and crop production [23]. Drought and cold stress lead to accumulation of the phytohormone abscisic acid (ABA), and it has been shown that application of ABA mimics stress conditions [24]. Therefore late embryogenesis abundant (LEA) proteins, which are induced by ABA and were shown to accumulate in vegetative organs during dehydration and low temperature stress [25,26], are good candidates to study adaptation. The LEA proteins are subdivided into seven groups based on their amino acid sequences as well as structural and functional features (e.g. size, hydrophilicity, or glycin content) [27]. In this study we analyze two types of LEA proteins: PLC30-15 [28] encoded by a drought and ABA-inducible dehydrin gene belonging to Group 2, and the ASRs, which belong to Group 7 [27]. Although the functional role of dehydrins still remains speculative, some observations suggest their involvement in abiotic stress tolerance. In Solanum tuberosum (potato) and S. sogarandinum, dehydrins are drought induced in apical parts and show an increased expression level correlated with cold tolerance in tubers and stems [29]. A previous study of the pLC30-15 dehydrin revealed that diversifying selection acted on its coding region in a wild tomato population from a dry environment [30]. The other genes used in this study are members of the ABA- and abiotic stress-induced Asr (ABA/water stress/ripening induced) gene family [31,32]. ASRs have several functions that help the plant dealing with stress: as monomers with a chaperon function in the cytoplasm [33] or as homo- and heterodimers with DREB (drought response element binding) proteins [34,35] with DNA-binding activity in the nucleus [36]. They also serve as transcription factors associated with the modulation of sugar transport activity [37-39]. Previous studies showed that over-expressed Asr genes in transgenic plants lead to higher drought and salt tolerance [40-42]. Using semi-quantitative RT-PCR in cultivated tomato (S. lycopersicum), it was demonstrated that Asr genes show differences in expression depending on the gene copy or the organ [43]. Analyzing different accessions of wild tomato using Northern Blots, it was shown that Asr1 and Asr4 are up-regulated in leaves of plants from humid environments after drought stress [44]. Other studies carried out in wild tomatoes revealed patterns consistent with local adaptation at Asr genes in populations that dwell in dry environments [45-47]. These findings make pLC30-15 and Asr genes interesting candidates for studying local adaptation at the gene expression level.

To understand their role in local adaptation, plants from their native environments are required [48]. For model organisms (e.g. A. thaliana, O. sativa, or Z. mays), an environmental context is not clear and/or cultivation caused reduced diversity due to bottlenecks and artificial selection. Investigating non-model organisms becomes more and more popular, but as they are mostly lacking sequenced genomes it is reasonable to study wild relatives of model organisms [49]. This has successfully been done in relatives of e.g. A. thaliana [50-53], Helianthus annuus (sunflower) [54,55], O. sativa [56], and S. lycopersicum [57]. The availability of cultivated tomato genomic resources, the recent divergence of the Solanum species, and their clear phenotypic distinction [58] make tomato species a popular plant system that is frequently used to study evolution [57,59]. Most Solanum sect. Lycopersicon species are native to western South America (Ecuador, Peru, and Chile), along the western and eastern Andean slopes [60]. This study focuses on two recently diverged wild tomato species that show differences in their ecological habitats and features: Solanum chilense and S. peruvianum. S. chilense is distributed from southern Peru to northern Chile where it inhabits arid plains and deserts [58]. It is known to be drought tolerant and can dwell in hyper-arid areas [57,58,61]. Furthermore, it shows a broad range in elevation from sea level up to 3,500 m and therefore experiences large temperature gradients during the year [62]. S. peruvianum is distributed from central Peru to northern Chile and inhabits a variety of habitats, from coastal deserts to river valleys [58].

At the level of populations, local adaptation can best be studied for organisms with restricted migration [63]. Using the coding sequences, previous population genetic analyses have provided evidence for local adaptation at Asr2, Asr4, and pLC30-15 [30,46,47]. Here we sequenced the regulatory regions of these genes from the same populations we had analyzed previously [30,47]. Therefore, we could investigate the evolutionary forces shaping the regulatory regions in direct comparison with the corresponding coding parts of the genes. In addition, we could identify conserved cis-acting elements. We also analyzed the expression pattern of Asr1, Asr2, Asr4, and pLC30-15 in S. chilense and S. peruvianum accessions that were sampled in close proximity to the populations used for the sequence analysis. We were able to determine differences in gene expression profiles (i.e. intensity and speed) and differences depending on the type of stress or the gene investigated.

Materials and Methods

Sequence analysis: Plant material and sequencing

We sequenced the promoter region of Asr2 (pAsr2), Asr4 (pAsr4), pLC30-15 (5’pLC), and also the downstream region of pLC30-15 (3’pLC). All genes are located on chromosome 4; genomic locations according to the SOL Genomics Network (http://solgenomics.net/) are as follows. Asr2: SL2.40ch04:56141779...56142589; Asr4: SL2.40ch04:56178656...56180338; pLC30-15: SL2.40ch04:63550865...63552237). Two populations from climatically different environments were sampled for each species (Tacna and Quicacha for S. chilense; Tarapaca and Canta for S. peruvianum). Five to seven individuals of each population were analyzed (Table S1 in File S1). A detailed description of these samples is provided in [47,59,64-66]. The Tarapaca sample was obtained from the Tomato Genetics Resource Center (TGRC) at the UC Davis (accession number LA2744). The other populations were sampled by T. Städler and T. Marczewski in May 2004 [66,67]. Relevant permits for the collection and import of samples from the Peruvian and Chilean Government were handled by T. Städler and T. Marczewski (as published in [66,67]) or the TGRC. Solanum ochranthum (TGRC accession LA2682) was used as outgroup. DNA extraction, PCR amplification, cloning, and sequencing were performed as described in [47]. All primers used in this project can be found in (Tables S2 and S3 in File S1). Sequence alignments are provided in Files S2-S5. Sequence data from this article have been deposited in the EMBL/GenBank Data Libraries under accession numbers HE612885-HE613033.

Nucleotide diversity analysis, neutrality tests, and haplotype diversity

We measured nucleotide diversity using Watterson’s θw and Tajima’s π implemented in the DnaSP v5 software [68]. θw is based on the number of segregating sites [69], and π on the average number of pairwise nucleotide differences among sequences in a sample [70]. We tested for deviations from the standard neutral model using Tajima’s D statistic and Fu & Li’s D in DnaSP. A significantly negative value of Tajima’s D indicates an excess of rare variants as expected under directional selection or population size expansion [71]. A significantly positive Tajima’s D value indicates an excess of intermediate-frequency variants as expected under balancing selection or in structured populations [71]. The same is true for Fu and Li’s D statistics [72]. Fay & Wu’s H was also calculated. A negative Fay and Wu’s H indicates an over-representation of high-frequency derived polymorphisms, which is expected under positive selection [73]. A positive Fay and Wu’s H indicates an over-representation of intermediate-frequency derived polymorphisms, which is the case if balancing selection was acting [73]. The haplotype test of Depaulis & Veuille [74] was used to assess haplotype diversity (Hd). All neutrality tests were performed using the option Number of Segregating Sites in DnaSP.

Motif search in non-coding regions

Motifs in the promoter region were searched using the program PlantCARE [75]. This program contains a database of cis-acting regulatory elements and allows for in silico analysis of promoter sequences. We limited our search to stress- and hormone-related motifs in conserved regions to provide more information on the kind of stresses that might trigger a response of those genes. We do not describe general transcription factor binding sites like the TATA box.

Gene expression analysis: Plant material, cultivation and replication

Due to restrictions by the Peruvian government only leaf material was sampled from populations used in the previous sequence analyses [30,47,59,66,67] and the promoter sequence analysis in this work. To perform the gene expression experiments, seeds of six accessions in close proximity to those previously sampled populations were obtained from the TGRC (Table 1). We tested five populations for drought stress and six populations for cold stress (Table 1). As these wild tomato species are outcrossing, we performed cuttings to obtain genetically identical replicates. Tomato seeds of the motherplants were treated with 2.7% NaOCl for 20 minutes to foster germination, and then kept on moistened filter paper at room temperature in the dark until they germinated. The tomato seedlings were then transferred to soil and put into the climate chamber at 22°C with a 16h/8h day/night cycle and 70% humidity. The motherplants were grown until they could provide material for 20-25 cuttings (approximately three months). The cuttings were treated with the Neudofix rooting enhancer (Neudorff, Emmerthal, Germany) and transferred to pots containing soil and vermiculite on top (to ensure nutrition and ventilation). The cuttings were grown for five weeks under the same conditions as the motherplants until they grew roots and three fresh leaves. The Asr genes and pLC30-15 were also sequenced in the motherplants to determine their haplotypes as described above. The experiment was designed as follows. Gene expression was measured at five time points after drought and cold stresses for five and six populations, respectively (Table 1). For each population, three biological and three technical replicates were used (for all time points).

AccessionaSpeciesbNearby populationcCollection sitebLatitude / Longitudeb   Altitudeb [m]Annual Precipitationd [mm]Precipitation wettest monthd [mm]Mean annual temperatured [°C]Stresses tested
LA1938 (QUI)S. chilenseQuicachaQuebrada Salsipuedes, Arequipa, Peru 15°41' S / 73°50' W1400 61 31 15.5drought + cold
LA1967 (TAC1)S. chilenseTacnaPachia, Tacna, Peru17°55' S / 70°09' W1000 15 5 16.7drought + cold
LA1969 (TAC2)S. chilenseTacnaEstique Pampa, Tacna, Peru17°32' S / 70°02' W3250 15 5 16.7cold
LA2744 (TAR1)S. peruvianumTarapacaSobraya, Tarapaca, Chile18°33' S / 70°09' W400 5 1 17.9drought + cold
LA2745 (TAR2)S. peruvianumTarapacaPan de Azucar, Tarapaca, Chile18°35' S / 69°56' W600 5 1 17.9drought + cold
LA3636 (CAN)S. peruvianumCantaCoayllo, Lima, Peru12°41' S / 76°24' WNo data available265 77 14.2drought + cold

Table 1. Location and habitat characteristics of the accessions used for the expression analysis.

aTomato Genetics Resource Center (TGRC) accession number
bAccording to TGRC database
cNearby populations sampled by T. Städler and T. Marczewski, 2004
dData extracted from WorldClim database (www.worldclim.org); precipitation driest month is always 0 mm
CSV
Download CSV

Stress treatment, RNA extraction and cDNA synthesis

Drought stress was applied by removing the tomato plants from the pots, carefully rinsing and drying their roots and transferring them into a climate chamber at 22°C (according to [43]). For the cold stress, the plants were transferred to a climate chamber at 4°C. Leaves were immediately frozen in liquid nitrogen at five timepoints: unstressed plants, one hour, three hours, six hours, and 24 hours after stress application. Total RNA was isolated using the RNeasy Plant Mini Kit (Qiagen GmbH, Hilden, Germany). DNA was removed using an on-column DNaseI digestion protocol. The RNA integrity was assessed by gel electrophoresis. A NanoDrop 1000 Spectrophotometer (Peqlab, Erlangen, Germany) was used to quantify the RNA and to assess its quality. Only samples with A260/A280 and A260/A230 values between 1.9 and 2.1 were used for further experiments. cDNA was synthesized from 1 µg total RNA using SuperScriptIII reverse transcriptase and RNase inhibitor RNaseOUT (both from Invitrogen, Carlsbad, CA, USA) using oligo(dT20) primers. The cDNA was treated with RNaseH (New England Biolabs, Ipswich, MA, USA) to remove remaining RNA.

Quantitative real-time PCR

Primers for the quantitative real-time PCR (in the following referred to as qPCR) were designed using NetPrimer (http://www.primierbiosoft.com/netprimer) and PrimerBLAST (http://www.ncbi.nlm.nih.gov). As Asr3 and Asr5 cannot be distinguished in their coding region [47] they were excluded from this study. qPCR was carried out using iQ SYBR green on a CFX thermocycler (both BioRad, Hercules, CA, USA). Expression of the target genes was normalized by two constitutively expressed reference genes: CT189 coding for a 40S ribosomal protein [65] and TIP4I which was shown to be a very stable reference gene in tomato [76]. As the efficiency was close to 100% for all runs we applied the 2-ΔΔCq method [77] to derive the relative expression quantity from the measured Cq-values. Quality control, reference gene stability, transformation to relative quantity, and normalization was carried out using the program qbasePLUS [78]. We performed the qPCR for one gene with all populations and timepoints in one run to rule out inter-run variation. To make the results between the different genes comparable, we performed an inter-run calibration in qbasePLUS. As this left us with more than one “true” timepoint 0 (= unstressed plants) for each gene we chose to display the results relative to the average of the whole run. We used the two-sample Wilcoxon (Mann-Whitney-U) test to determine whether differences in relative expression between stressed and unstressed plants were significant [79]. Importantly, we only compare timepoints within populations to make inferences on the time of gene up-regulation and the peak of expression. As we observe differences in relative expression between the populations at timepoint 0, we do not make comparisons between populations at the same timepoint as this could be misleading.

Results

Incomplete selective sweep in Asr2 promoter region in the Quicacha population

We sequenced ~1,500 bp upstream of Asr2, Asr4, pLC30-15 as this region should contain most cis-regulatory regions (i.e. transcription factor binding sites). Additionally, we sequenced the downstream region of pLC30-15 (~2,300 bp) to determine how far the signature of positive selection at this gene found by [30] extends. However, we cannot rule out other cis- or trans-regulatory regions. We wanted to investigate the evolutionary forces acting on regulatory regions of genes involved in stress response. pAsr2 shows a low nucleotide diversity, especially at Quicacha, compared to the Asr2 coding region (Table 2, Figure S1a in File S1). Haplotype diversity is also very low in the Quicacha population (Table 3) and Tajima’s D and Fu and Li’s D are significantly negative (Table 4, Figure S1b in File S1). Indeed, we found only two pAsr2 haplotypes at Quicacha that were rather similar to each other, where the minor allele occurred only once (Figure S3 in File S1). This indicates that positive directional selection has been acting in this population at pAsr2, causing an incomplete selective sweep [66,80].

πpAsr2Asr2apAsr4Asr4a5' pLCpLCb3' pLC
Quicacha0.0040.0160.0220.0090.0300.0140.023
Tacna0.0150.0200.0260.0150.0280.0120.016
Canta0.0160.0200.0320.0190.0440.0160.027
Tarapaca0.0150.0220.0310.0210.0430.0120.022
θW
Quicacha0.0060.0140.0210.0110.0260.0100.023
Tacna0.0130.0220.0260.0220.0340.0130.017
Canta0.0200.0200.0350.0210.0440.0160.031
Tarapaca0.0180.0210.0300.0200.0410.0120.023

Table 2. Nucleotide diversity of pAsr2, pAsr4, 5’pLC, 3’pLC, and their corresponding genes.

aFrom [47]
bFrom [30]
CSV
Download CSV
HdpAsr2Asr2apAsr4Asr4a5' pLCpLCb3' pLC
Quicacha0.2860.8000.9330.6000.6670.7110.889
Tacna0.9330.9780.9820.9821.0000.9780.982
Canta0.9450.9331.0000.9561.0000.9780.972
Tarapaca0.9640.9560.9050.9330.8890.8220.889

Table 3. Haplotype diversity of pAsr2, pAsr4, 5’pLC, 3’pLC, and their corresponding genes.

aFrom [47]
bFrom [30]
CSV
Download CSV
Tajima's DpAsr2cAsr2apAsr4Asr4a5' pLCpLCb3' pLC
Quicacha-1.688*0.6020.356-0.5110.8872.342*0.022
Tacna1.122-0.3720.078-1.429-0.811-0.358-0.499
Canta-0.936-0.121-0.432-0.5480.008-0.339-0.700
Tarapaca-0.7830.3560.1700.0290.1370.010-0.144
Fu and Li's D
Quicacha-1.791**-0.1210.8830.3821.853**1.587**-0.432
Tacna0.6610.031-0.230-2.326*-1.220-0.3840.281
Canta-1.423-0.590-1.179-0.982-0.916-0.673-1.497
Tarapaca-0.7840.225-0.214-0.2230.1800.119-0.401
Fay and Wu's H
QuicachaNA-0.927-3.556-8.709-9.190-1.067-2.667
TacnaNA5.6263.200-5.127-7.7271.511-28.673
CantaNA0.4443.7781.8677.9110.9788.806
TarapacaNA0.0005.1430.889-3.1672.3111.511

Table 4. Results of the neutrality tests for pAsr2, pAsr4, 5’pLC, 3’pLC, and their corresponding genes.

aFrom [47]
bFrom [30]
cNo outgroup available. Fu and Li’s D* without outgroup was calculated instead.
NA Not applicable
* P<0.05, ** P<0.01 (significant results are in bold)
CSV
Download CSV

The evolution of the Asr4 and the pLC30-15 regulatory regions

Compared to the Asr4 coding region, where evidence for local adaptation has been shown in a population from a dry environment [47], nucleotide diversity is higher upstream of the gene (Table 2). Additionally, the low haplotype diversity observed at Asr4 in the Quicacha population increases at pAsr4 to almost 1 (Table 3), and no deviation from neutrality can be observed (Table 4). This indicates that the forces acting on the Asr4 gene (directional selection) are weaker in the promoter region. The relatively low level of haplotype diversity detected at pLC30-15 in Quicacha can also be found at 5’pLC (Table 3, Figure S4 in File S1). Additionally, Fay and Wu’s H becomes very negative in both Quicacha and Tacna (Table 4), indicating positive (diversifying) selection. Nucleotide diversity is generally higher at 3’pLC than at the pLC30-15 gene except in the Tacna population (Table 2, Figure S2a in File S1). In addition, Fay and Wu’s H becomes extremely negative at the downstream region of the pLC30-15 gene, indicating positive selection (Table 4, Figure S2b in File S1).

Different types of motifs in the regulatory regions of Asr2, Asr4 and pLC30-15

We analyzed the regulatory regions of Asr2, Asr4, and pLC30-15 in silico to identify cis-acting elements (short motifs of 4-10 bases). Here we describe only those motifs related to hormone and stress response, and motifs that lie in conserved regions (i.e. without polymorphism) in the alignment of all sequences of both species (see Table S4 in File S1 for additional information of the described motifs). At pAsr2 we detected a motif conserved in both species involved in salicylic acid responsiveness (TCA-element). We also discovered one motif conserved in S. chilense involved in ethylene (ERE). This motif (conserved in S. peruvianum) as well as an abscisic acid responsive element (ABRE) was also found at pAsr4. In addition, pAsr4 contains a conserved auxin responsive element (Aux-RR-core). At 5’pLC we found five conserved ABRE motifs and a motif involved in methyl jasmonate responsiveness (CGTCA-motif); at 3’pLC conserved ABRE and CGTCA-motifs were detected as well as an AuxRR-core.

Conserved stress-related regulatory elements at pAsr2 are involved in anaerobic induction (ARE), drought responsiveness (MBS), and general stress and defense response (TC-rich repeats). At pAsr4 a conserved stress related motif is ARE. TC-rich repeats and MBS motifs are conserved at 5’pLC and 3’pLC. 3’pLC also contains motifs involved in low-temperature responsiveness (LTR) and heat stress responsiveness (HSE).

Expression patterns of Asr1 and Asr2

We analyzed the expression patterns of Asr1, Asr2, Asr4, and pLC30-15 in different populations using a time-course experiment after exposing wild tomato plants to cold and drought stress. In general, all genes appear to be more strongly induced in S. chilense than in S. peruvianum and respond more strongly to water deficit than to cold stress (Figures 1-4). After application of drought stress, Asr1 is induced after 1-3h in all S. chilense and S. peruvianum accessions (Figure 1a+b). After cold stress, Asr1 is induced after 1-6h but at a relatively low level in all S. chilense and S. peruvianum accessions (Figure 1c+d).

thumbnail
Figure 1. Gene expression of Asr1 after application of drought and cold stress.

Expression is displayed relative to the average of the whole qPCR run in unstressed plants, 1h, 3h, 6h, and 24h after stress application. (A) The following accessions of S. chilense were measured after drought stress (red bars): LA1938 (QUI) from a dry environment and LA1967 (TAC) from a hyperarid area. (C) The following accessions of S. chilense were measured after cold stress (blue bars): LA1938 (QUI) from a dry environment and LA1967 (TAC1) from a hyperarid area and LA1969 (TAC2) from a very dry environment and high altitude. The following accessions of S. peruvianum were measured after (B) drought (red bars) and (D) cold stress (blue bars): LA2744 (TAR1) and LA2745 (TAR2) from a dry environment and LA3636 (CAN) from a humid environment. Vertical lines at bar charts indicate the standard error, asterisks above bar charts indicate significant over-expression compared to the unstressed control (*P<0.05; **P<0.01), arrows above bar charts indicate significant over-expression (P<0.01) compared to the previous timepoint – meaning the transcript level is increasing.

https://doi.org/10.1371/journal.pone.0078182.g001

thumbnail
Figure 2. Gene expression of Asr2 after application of drought and cold stress.

(for explanation see Figure 1).

https://doi.org/10.1371/journal.pone.0078182.g002

thumbnail
Figure 3. Gene expression of Asr4 after application of drought and cold stress.

(for explanation see Figure 1).

https://doi.org/10.1371/journal.pone.0078182.g003

thumbnail
Figure 4. Gene expression of pLC30-15 after application of drought and cold stress.

(for explanation see Figure 1).

https://doi.org/10.1371/journal.pone.0078182.g004

Relative Asr2 expression is generally quite low compared to the other genes studied here. When applying drought stress to S. chilense populations, Asr2 is significantly up-regulated after 1h in Tacna and 6h in Quicacha (Figure 2a). In S. peruvianum, we observe slightly different Asr2 expression patterns in accessions from the same environment: transcription is induced after 1h in one Tarapaca accession (LA2744) and only after 24h in the other Tarapaca accession (LA2745) and Canta (Figure 2b). After application of cold stress in S. chilense, Asr2 is significantly up-regulated after 1h in Quicacha and the Tacna accession from a high altitude (LA1969; 3,250m) or after 3h in the Tacna accession from a lower altitude (LA1967; Figure 2c). In S. peruvianum, LA2745 shows the highest and fastest (1h) induction of Asr2 (Figure 2d). LA2744 and the Canta accession are induced more slowly (3h and 24h, respectively; Figure 2d).

Faster induction of Asr4 transcription in a population from a dry environment

In the Quicacha accession of S. chilense, Asr4 is drought induced after 3h and very highly expressed after 24h (Figure 3a). In the Tacna accession from a very dry environment, on the other hand, Asr4 is already significantly up-regulated after 1h and has its expression peak after 3h (Figure 3a). Based on DNA sequence variation data this population was found to undergo local adaptation and to exhibit signatures of positive selection at the Asr4 locus [47]. The Tacna accession used in the expression experiments is homozygous for the favored predominant haplotype described in [47]. In S. peruvianum, the two Tarapaca accessions (both from a similar environment) differ in their expression patterns. In the accession LA2745, Asr4 is drought induced after 1h and reaches its maximal expression level after 6h (Figure 3b). LA2744, on the other hand, shows a constant increase of Asr4 transcripts until timepoint 24h (Figure 3b). The Canta accession displays a significant over-expression of Asr4 after 6h and 24h (Figure 3b).

Asr4 induction after cold stress is much lower. All S. chilense accessions show a fast induction (1h) of Asr4 transcripts (Figure 3c). However, the transcript level seems to be higher in the Tacna accession from a high altitude (LA1969) compared to the one from a lower altitude (LA1967; <1,000m) and the accession from Quicacha. In S. peruvianum, Asr4 induction is much slower: after 3h in Canta and 6h in Tarapaca (Figure 3d).

High expression levels of pLC30-15 after drought stress

After application of drought stress, pLC30-15 is induced fast (after 1h in all accessions) and reaches its peak of transcription after 3-6h – except for the Quicacha and Canta accessions where the expression level increases until 24h (Figure 4a,b). Transcript levels appear to be higher in S. chilense than in S. peruvianum after drought stress, but to a lesser degree than for the Asr genes (Figure 4a,b). After cold stress, pLC30-15 has a lower transcript level in both species (Figure 4c,d). pLC30-15 is significantly up-regulated after 1h in all S. chilense accessions and expression keeps increasing until 24h (Figure 4c). In S. peruvianum, we can again observe differences in expression in accessions from the same environment: pLC30-15 is induced after 1h in LA2744 but no induction can be detected in accession LA2745 (both TarapacaFigure 4d). In Canta, pLC30-15 is induced after 3h and increases until 24h (Figure 4d).

Discussion

We analyzed the regulatory regions of stress-responsive genes in wild tomato populations from different environments and compared them to their corresponding coding regions, which were previously studied [30,47]. Our most salient observations are as follows. Sequence analyses suggest that the Asr2 promoter region and the downstream region of pLC15-30 in S. chilense populations from dry environments have been under positive selection. The gene expression experiments suggest that, in general, the genes show a higher induction in S. chilense than in S. peruvianum and respond more strongly to water deficit than to cold stress. In particular, we found that pLC30-15 and Asr4 are highly drought induced in the S. chilense population from Tacna (a very dry environment). As these genes also exhibit signatures of positive selection in the coding region [30,47] they may therefore be of potential interest for further functional studies of adaptation. Since we did not perform statistical analysis between populations, however, these observed trends are suggestive and further experiments are needed to verify them. In the following, we discuss these findings in more detail.

Sequence variation of regulatory regions of stress-responsive genes

Studying cis-regulatory elements has been of great interest over the last years as it has been suggested that phenotypic changes also result from variation in these regions rather than in coding regions [2,81,82]. A relatively straightforward way to investigate this is to first analyze sequence variation and search for specific motifs in promoter regions [83]. When analyzing the Asr2 promoter region, it is quite remarkable how low nucleotide diversity is especially in the Quicacha population. The observed polymorphism pattern suggests that a (incomplete) selective sweep eliminated nucleotide diversity at pAsr2 and increased the frequency of one favored haplotype. This may indicate an important function of this region. Similarly, the downstream region of pLC30-15 shows low nucleotide diversity and patterns consistent with positive selection in the Tacna population, which may also be functionally significant. Unlike pAsr2 and 3’pLC, pAsr4 and 5’pLC are more polymorphic than their corresponding coding regions. The patterns consistent with local adaptation at Asr4 in the Tacna population described before [47] disappear at pAsr4 and neutrality tests show no deviation from a standard neutral scenario. Also, the haplotype structure observed in Quicacha [47] is not present at pAsr4. This shows that the evidence for local adaptation described before is limited to the Asr4 coding regions. On the other hand, the haplotype structure and patterns of positive selection at the pLC30-15 locus described in [30] in Quicacha remain at 5’pLC, suggesting that the acting selective forces are not limited to the gene.

Although we found evidence for positive selection in the coding regions of Asr2, Asr4 and pLC30-15 and also in some of their regulatory regions, it is difficult to establish a relationship between the sequence evolution of the genes and their expression profiles. Except for the possible effects of trans-regulatory elements, reasons for this may be that more samples from different environments, larger sample sizes and more genes need to be investigated. Nonetheless, our approach may prove useful as an initial step towards determining whether these genes are involved in the adaptation of wild tomatoes to abiotic factors such as drought and cold. Indeed, an encouraging sign might be that our motif scan analysis shows conservative regulatory elements that are involved in drought, cold, heat stress and general stress responses at pAsr2, 5’pLC and 3’pLC. Further evidence about the possible relationship between sequence and gene expression variation is discussed below, after reviewing the results from related studies.

Environment-specific gene expression regulation of stress-responsive genes and its possible relationship to sequence variation

Asr genes have already been described to be induced by desiccation in several plant species, e.g. Solanum chacoense (wild potato) [84], Pinus taeda (loblolly pine) [85], Lilium longifolium (lily) [86], Ginkgo biloba [87], and O. sativa [88], but also by cold in S. tuberosum [89]. Asr genes have been shown to be very variable in their expression kinetics. They show differences in organ-specific expression [43,85] and different patterns depending on the gene copy [44,88] and applied stress [87,89]. Philippe et al. [88] even demonstrate differential expression patterns of rice Asr genes depending on the cultivar. Such variability in gene expression between accessions was also described for cold-responsive genes in wild tomato [18]. Similarly, although they share the same environment, the S. peruvianum accessions from Tarapaca (LA2744 and LA2745) seem to show different expression patterns for Asr2, Asr4 and pLC30-15. Therefore, differences in gene expression may not necessarily be adaptive. It has been demonstrated that stress-related genes have a more variable expression than housekeeping genes [90,91] and a higher expression divergence of duplicated genes in A. thaliana [92]. In tomato, we detected different expression patterns between populations from similar environments only if the genes were in general lowly expressed. This is in accordance with the findings of Carey et al. [93], who found that transcriptional noise decreases as the expression level increases.

Although we cannot establish a clear connection between gene expression variation and patterns of diversity in the regulatory regions or the coding regions analyzed in previous studies [30,45-47], we discovered interesting trends. First, we suggest that the Asr genes and pLC30-15 are more strongly induced by water deficit than by cold, indicating that they play a more important role in drought response than in cold response in wild tomato species. In accordance with the results by Carey et al. [93] discussed above, the expression patterns of the Asr genes and pLC30-15 appear to be more noisy after cold stress. Second, Asr1 seems to show a similar expression pattern after drought stress in all S. chilense and S. peruvianum populations, i.e. induction occurs after 1-3h and relative expression increases until 24h, while the expression level is lower and more noisy after cold stress. Asr1 homologs have been shown to be conserved in wild tomato and other plant species and seem to act as housekeeping genes [44,47]. Our results suggest a concordant conserved expression pattern, highlighting the importance of Asr1 for basic functions in the plants. Third, Asr2 expression appears to be quite low compared to the other candidate genes. This might indicate that Asr2 does not play a major role in drought and cold response, but rather in other stress conditions. Another possible explanation for the relatively low expression of Asr2 is that it is predominantly expressed in other organs than leaves. Indeed, Maskin et al. [43] found Asr2 to be up-regulated in roots – but not in leaves – of cultivated tomato after drought stress. As we do not discover a high expression in leaves, organ-specific expression of Asr2 might also occur in wild tomato species. However, more tissues should be tested to validate these findings.

Finally, under water deficit pLC30-15 and Asr4 are more quickly induced in the accession from Tacna (hyperarid habitat) than in the other S. chilense accession from Quicacha (less dry habitat). However, we also observe a down-regulation of Asr4 in the accession from Tacna after 3h. This could explain previous findings [44], in which Asr4 expression could not be detected after 24h in drought-stressed wild tomato plants from a dry environment. Induction and down-regulation are faster in populations from dry habitats and the transcript was not sufficiently abundant to be detected. In the Tacna population, Fischer et al. [47] described a predominant Asr4 haplotype, which is absent in other S. chilense populations. Interestingly, the accession tested here is homozygous for this haplotype which further highlights the abundance of this haplotype in the Tacna population. In addition, after cold stress Asr4 seems to be more strongly induced in the Tacna accession from a high altitude compared to the other S. chilense accessions. As for the other genes studied here, however, Asr4 expression appears to be in general much lower and therefore tends to be noisier after cold treatment.

Conclusions

Wild relatives of crop species have many advantages that make them interesting resources for studying plant evolution. One of them is that they are sampled from the ecological context they evolved in. We analyzed the expression variation of candidate genes in natural wild tomato populations. Such a rare study gives insights into natural variation in gene expression and can provide good candidates for improving plant tolerance to abiotic stresses. We found Asr4 to be an interesting candidate, in accordance with a previous study [47]. Our observations suggest that both Asr2 and Asr4 as well as pLC30-15 are induced by abiotic stresses, particularly by drought. The present study, as well as some others carried out in wild tomatoes [30,45-47], indicate that a candidate gene approach is efficient for detecting evidence for local adaptation to abiotic stresses and that wild tomato species constitute a valuable genetic resource for genes conferring resistance to abiotic stress. The genome of cultivated tomato became recently available [94]. Therefore, the evolution of regulatory elements can now be analyzed much more comprehensively, as has been done in Arabidopsis [95]. Finally, our results that positive selection occurs more often in local S. chilense populations and gene expression responses appear to be generally faster and stronger in this species seem to support the previous conclusions that S. chilense shows more evidence of local adaptation to drought and temperature stress than S. peruvianum [30,47,96].

Supporting Information

File S1.

Contains additional Figures and Tables: Table S1 Numbers of sequenced haplotypes. Table S2 Primer sequences and amplification details for PCR of pAsr2, pAsr4, 5’pLC, and 3’pLC. Table S3 Primer sequences and amplification details for the qPCR of the Asr genes, pLC30-15, and the reference genes. Table S4 Summary of function and sequences of motifs found at pAsr2, pAsr4, 5'pLC, and 3'pLC using PlantCARE. Figure S1: Sliding window analysis of (A) π and (B) Tajima’s D for the Tacna (red) and the Quicacha (green) populations over pAsr2. The x-axis indicates the location relative to the start codon of the Asr2 gene; purple boxes indicate regulatory motifs. Figure S2: Sliding window analysis of (A) π and (B) Fay & Wu’s H for the Tacna (red) and the Quicacha (green) populations over 3’pLC. The x-axis indicates the location relative to the stop codon of the pLC30-15 gene; purple boxes indicate regulatory motifs. Figure S3: pAsr2 haplotypes for the Tacna and Quicacha populations. Only polymorphic sites are shown. The number behind the sequences indicates the frequency of each haplotype. Figure S4: 5’pLC haplotypes for the Tacna and Quicacha populations. Only polymorphic sites are shown. The number behind the sequences indicates the frequency of each haplotype.

https://doi.org/10.1371/journal.pone.0078182.s001

(PDF)

Acknowledgments

We thank Hilde Lainer for technical assistance, Susi Voigt for valuable suggestions on the gene expression analysis, and Lu Zhang for help with the expression experiments. We are also grateful to Létizia Camus-Kulandaivelu, Miri Linnenbrink and three reviewers for valuable comments that improved the presentation of this paper.

Author Contributions

Conceived and designed the experiments: IF WS MM. Performed the experiments: IF KS MM. Analyzed the data: IF KS MM. Wrote the manuscript: IF WS MM.

References

  1. 1. Fisher RA (1930) The genetical theory of natural selection. Oxford, UK: Oxford University Press.
  2. 2. Wray GA (2007) The evolutionary significance of cis-regulatory mutations. Nat Rev Genet 8: 206-216. doi:10.1038/nrg2063. PubMed: 17304246.
  3. 3. Wilson AC, Maxson LR, Sarich VM (1974) Two types of molecular evolution - evidence from studies of interspecific hybridization. Proc Natl Acad Sci U_S_A 71: 2843-2847. doi:10.1073/pnas.71.7.2843. PubMed: 4212492.
  4. 4. King MC, Wilson AC (1975) Evolution at two levels in humans and chimpanzees. Science 188: 107-116. doi:10.1126/science.1090005. PubMed: 1090005.
  5. 5. Ferea TL, Botstein D, Brown PO, Rosenzweig RF (1999) Systematic changes in gene expression patterns following adaptive evolution in yeast. Proc Natl Acad Sci U_S_A 96: 9721-9726. doi:10.1073/pnas.96.17.9721. PubMed: 10449761.
  6. 6. Cooper TF, Rozen DE, Lenski RE (2003) Parallel changes in gene expression after 20,000 generations of evolution in Escherichia coli. Proc Natl Acad Sci U_S_A 100: 1072-1077. doi:10.1073/pnas.0334340100. PubMed: 12538876.
  7. 7. Doebley J, Stec A, Hubbard L (1997) The evolution of apical dominance in maize. Nature 386: 485-488. doi:10.1038/386485a0. PubMed: 9087405.
  8. 8. Wang H, Nussbaum-Wagler T, Li B, Zhao Q, Vigouroux Y et al. (2005) The origin of the naked grains of maize. Nature 436: 714-719. doi:10.1038/nature03863. PubMed: 16079849.
  9. 9. Konishi S, Izawa T, Lin SY, Ebana K, Fukuta Y et al. (2006) An SNP caused loss of seed shattering during rice domestication. Science 312: 1392-1396. doi:10.1126/science.1126410. PubMed: 16614172.
  10. 10. Hutter S, Saminadin-Peter SS, Stephan W, Parsch J (2008) Gene expression variation in African and European populations of Drosophila melanogaster. Genome Biol 9: R12. doi:10.1186/gb-2008-9-1-r12. PubMed: 18208589.
  11. 11. Müller L, Hutter S, Stamboliyska R, Saminadin-Peter SS, Stephan W et al. (2011) Population transcriptomics of Drosophila melanogaster females. BMC Genomics 12: 81. doi:10.1186/1471-2164-12-81. PubMed: 21276238.
  12. 12. Matzkin LM (2012) Population transcriptomics of cactus host shifts in Drosophila mojavensis. Mol Ecol 21: 2428-2439. doi:10.1111/j.1365-294X.2012.05549.x. PubMed: 22512269.
  13. 13. Hodgins KA, Lai Z, Nurkowski K, Huang J, Rieseberg LH (2013) The molecular basis of invasiveness: differences in gene expression of native and introduced common ragweed (Ambrosia artemisiifolia) in stressful and benign environments. Mol Ecol 22: 2496-2510. doi:10.1111/mec.12179. PubMed: 23294156.
  14. 14. Guggisberg A, Lai Z, Huang J, Rieseberg LH (2013) Transcriptome divergence between introduced and native populations of Canada thistle, Cirsium arvense. New Phytol 199: 595-608. doi:10.1111/nph.12258. PubMed: 23586922.
  15. 15. Catalán A, Hutter S, Parsch J (2012) Population and sex differences in Drosophila melanogaster brain gene expression. BMC Genomics 13: 654. doi:10.1186/1471-2164-13-654. PubMed: 23170910.
  16. 16. Smith G, Fang Y, Liu X, Kenny J, Cossins AR et al. (2013) Transcriptome-wide expression variation associated with environmental plasticity and mating success in cactophilic Drosophila mojavensis. Evolution 67: 1950-1963. doi:10.1111/evo.12082. PubMed: 23815652.
  17. 17. Guo S, Zhang J, Sun H, Salse J, Lucas WJ et al. (2013) The draft genome of watermelon (Citrullus lanatus) and resequencing of 20 diverse accessions. Nat Genet 45: 51-58. PubMed: 23179023.
  18. 18. Mboup M, Fischer I, Lainer H, Stephan W (2012) Trans-species polymorphism and allele-specific expression in the CBF gene family of wild tomatoes. Mol Biol Evol 29: 3641-3652. doi:10.1093/molbev/mss176. PubMed: 22787283.
  19. 19. Baker RL, Hileman LC, Diggle PK (2012) Patterns of shoot architecture in locally adapted populations are linked to intraspecific differences in gene regulation. New Phytol 196: 271-281. doi:10.1111/j.1469-8137.2012.04245.x. PubMed: 22882227.
  20. 20. Fay JC, Wittkopp PJ (2008) Evaluating the role of natural selection in the evolution of gene regulation. Heredity 100: 191-199. doi:10.1038/sj.hdy.6801000. PubMed: 17519966.
  21. 21. Schaefke B, Emerson JJ, Wang T-Y, Lu M-YJ, Hsieh L-C et al. (2013) Inheritance of gene expression level and selective constraints on trans- and cis-regulatory changes in yeast. Mol Biol Evol, 30: 2121–33. doi:10.1093/molbev/mst1114. PubMed: 23793114.
  22. 22. Li C-M, Tzeng J-N, Sung H-M (2012) Effects of cis and trans regulatory variations on the expression divergence of heat shock response genes between yeast strains. Gene 506: 93-97. doi:10.1016/j.gene.2012.06.034. PubMed: 22759523.
  23. 23. Yáñez M, Cáceres S, Orellana S, Bastías A, Verdugo I et al. (2009) An abiotic stress-responsive bZIP transcription factor from wild and cultivated tomatoes regulates stress-related genes. Plant Cell Rep 28: 1497-1507. doi:10.1007/s00299-009-0749-4. PubMed: 19652975.
  24. 24. Mahajan S, Tuteja N (2005) Cold, salinity and drought stresses: an overview. Arch Biochem Biophys 444: 139-158. doi:10.1016/j.abb.2005.10.018. PubMed: 16309626.
  25. 25. Ingram J, Bartels D (1996) The molecular basis of dehydration tolerance in plants. Annu Rev Plant Physiol Plant Mol Ecol 47: 377-403.
  26. 26. Bray EA (1997) Plant responses to water deficit. Trends Plant Sci 2: 48-54. doi:10.5363/tits.2.10_48.
  27. 27. Battaglia M, Olvera-Carrillo Y, Garciarrubio A, Campos F, Covarrubias AA (2008) The enigmatic LEA proteins and other hydrophilins. Plant Physiol 148: 6-24. doi:10.1104/pp.108.120725. PubMed: 18772351.
  28. 28. Chen RD, Campeau N, Greer AF, Bellemare G, Tabaeizadeh Z (1993) Sequence of a novel abscisic acid-induced and drought-induced cDNA from wild tomato (Lycopersicon chilense). Plant Physiol 103: 301. doi:10.1104/pp.103.1.301. PubMed: 8208856.
  29. 29. Rorat T, Szabala BM, Grygorowicz WJ, Wojtowicz B, Yin Z et al. (2006) Expression of SK3-type dehydrin in transporting organs is associated with cold acclimation in Solanum species. Planta 224: 205-221. doi:10.1007/s00425-005-0200-1. PubMed: 16404580.
  30. 30. Xia H, Camus-Kulandaivelu L, Stephan W, Tellier A, Zhang Z (2010) Nucleotide diversity patterns of local adaptation at drought-related candidate genes in wild tomatoes. Mol Ecol 19: 4144–4154. doi:10.1111/j.1365-294X.2010.04762.x. PubMed: 20831645.
  31. 31. Iusem ND, Bartholomew DM, Hitz WD, Scolnik PA (1993) Tomato (Lycopersicon esculentum) transcript induced by water deficit and ripening. Plant Physiol 102: 1353-1354. doi:10.1104/pp.102.4.1353. PubMed: 8278555.
  32. 32. Rossi M, Iusem ND (1994) Tomato (Lycopersicon esculentum) genomic clone homologous to a gene encoding an abscisic acid induced protein. Plant Physiol 104: 1073-1074. doi:10.1104/pp.104.3.1073. PubMed: 8165244.
  33. 33. Konrad Z, Bar-Zvi D (2008) Synergism between the chaperone-like activity of the stress regulated ASR1 protein and the osmolyte glycine-betaine. Planta 227: 1213-1219. doi:10.1007/s00425-008-0693-5. PubMed: 18270732.
  34. 34. Maskin L, Frankel N, Gudesblat G, Demergasso MJ, Pietrasanta LI et al. (2007) Dimerization and DNA-binding of ASR1, a small hydrophilic protein abundant in plant tissues suffering from water loss. Biochem Biophys Res Commun 352: 831-835. doi:10.1016/j.bbrc.2006.11.115. PubMed: 17157822.
  35. 35. Saumonneau A, Agasse A, Bidoyen MT, Lallemand M, Cantereau A et al. (2008) Interaction of grape ASR proteins with a DREB transcription factor in the nucleus. FEBS Lett 582: 3281-3287. doi:10.1016/j.febslet.2008.09.015. PubMed: 18804467.
  36. 36. Kalifa Y, Gilad A, Konrad Z, Zaccai M, Scolnik PA et al. (2004) The water- and salt-stress-regulated Asr1 (abscisic acid stress ripening) gene encodes a zinc-dependent DNA-binding protein. Biochem J 381: 373-378. doi:10.1042/BJ20031800. PubMed: 15101820.
  37. 37. Carrari F, Fernie AR, Iusem ND (2004) Heard it through the grapevine? ABA and sugar cross-talk: the ASR story. Trends Plant Sci 9: 57-59. doi:10.1016/j.tplants.2003.12.004. PubMed: 15106586.
  38. 38. Frankel N, Nunes-Nesi A, Balbo I, Mazuch J, Centeno D et al. (2007) ci21A/Asr1 expression influences glucose accumulation in potato tubers. Plant Mol Biol 63: 719-730. doi:10.1007/s11103-006-9120-0. PubMed: 17211513.
  39. 39. Maskin L, Maldonado S, Iusem ND (2008) Tomato leaf spatial expression of stress-induced Asr genes. Mol Biol Rep 35: 501-505. doi:10.1007/s11033-007-9114-2. PubMed: 17602312.
  40. 40. Jeanneau M, Gerentes D, Foueillassar X, Zivy M, Vidal J et al. (2002) Improvement of drought tolerance in maize: towards the functional validation of the Zm-Asr1 gene and increase of water use efficiency by over-expressing C4-PEPC. Biochimie 84: 1127-1135. doi:10.1016/S0300-9084(02)00024-X. PubMed: 12595141.
  41. 41. Kalifa Y, Perlson E, Gilad A, Konrad Z, Scolnik PA et al. (2004) Over-expression of the water and salt stress-regulated Asr1 gene confers an increased salt tolerance. Plant Cell Environ 27: 1459-1468. doi:10.1111/j.1365-3040.2004.01251.x.
  42. 42. Yang C-Y, Chen Y-C, Jauh GY, Wang C-S (2005) A lily ASR protein involves abscisic acid signaling and confers drought and salt resistance in Arabidopsis. Plant Physiol 139: 836-846. doi:10.1104/pp.105.065458. PubMed: 16169963.
  43. 43. Maskin L, Gudesblat GE, Moreno JE, Carrari FO, Frankel N et al. (2001) Differential expression of the members of the Asr gene family in tomato (Lycopersicon esculentum). Plant Sci 161: 739-746. doi:10.1016/S0168-9452(01)00464-2.
  44. 44. Frankel N, Carrari F, Hasson E, Iusem ND (2006) Evolutionary history of the Asr gene family. Gene 378: 74-83. doi:10.1016/j.gene.2006.05.010. PubMed: 16822623.
  45. 45. Frankel N, Hasson E, Iusem ND, Rossi MS (2003) Adaptive evolution of the water stress-induced gene Asr2 in Lycopersicon species dwelling in arid habitats. Mol Biol Evol 20: 1955-1962. doi:10.1093/molbev/msg214. PubMed: 12949146.
  46. 46. Giombini MI, Frankel N, Iusem ND, Hasson E (2009) Nucleotide polymorphism in the drought responsive gene Asr2 in wild populations of tomato. Genetica 136: 13-25. doi:10.1007/s10709-008-9295-1. PubMed: 18636230.
  47. 47. Fischer I, Camus-Kulandaivelu L, Allal F, Stephan W (2011) Adaptation to drought in two wild tomato species: the evolution of the Asr gene family. New Phytol 190: 1032-1044. doi:10.1111/j.1469-8137.2011.03648.x. PubMed: 21323928.
  48. 48. Anderson JT, Willis JH, Mitchell-Olds T (2011) Evolutionary genetics of plant adaptation. Trends Genet 27: 258-266. doi:10.1016/j.tig.2011.04.001. PubMed: 21550682.
  49. 49. Song BH, Mitchell-Olds T (2011) Evolutionary and ecological genomics of non-model plants. J Syst Evol 49: 17-24. doi:10.1111/j.1759-6831.2010.00111.x. PubMed: 21394233.
  50. 50. Riihimäki M, Podolsky R, Kuittinen H, Koelewijn H, Savolainen O (2005) Studying genetics of adaptive variation in model organisms: flowering time variation in Arabidopsis lyrata. Genetica 123: 63-74. doi:10.1007/s10709-003-2711-7. PubMed: 15881681.
  51. 51. Turner TL, Bourne EC, Von Wettberg EJ, Hu TT, Nuzhdin SV (2010) Population resequencing reveals local adaptation of Arabidopsis lyrata to serpentine soils. Nat Genet 42: 260-263. doi:10.1038/ng.515. PubMed: 20101244.
  52. 52. Leinonen PH, Remington DL, Savolainen O (2011) Local adaptation, phenotypic differentiation, and hybrid fitness in diverged natural populations of Arabidopsis lyrata. Evolution 65: 90-107. doi:10.1111/j.1558-5646.2010.01119.x. PubMed: 20812972.
  53. 53. Knight CA, Vogel H, Kroymann J, Shumate A, Witsenboer H et al. (2006) Expression profiling and local adaptation of Boechera holboellii populations for water use efficiency across a naturally occurring water stress gradient. Mol Ecol 15: 1229-1237. doi:10.1111/j.1365-294X.2006.02818.x. PubMed: 16626450.
  54. 54. Kane NC, Rieseberg LH (2007) Selective sweeps reveal candidate genes for adaptation to drought and salt tolerance in common sunflower, Helianthus annuus. Genetics 175: 1823-1834. doi:10.1534/genetics.106.067728. PubMed: 17237516.
  55. 55. Kane NC, Rieseberg LH (2008) Genetics and evolution of weedy Helianthus annuus populations: adaptation of an agricultural weed. Mol Ecol 17: 384-394. doi:10.1111/j.1365-294X.2007.03467.x. PubMed: 17725567.
  56. 56. Grillo MA, Li C, Fowlkes AM, Briggeman TM, Zhou A et al. (2009) Genetic architecture for the adaptive origin of annual wild rice, Oryza nivara. Evolution 63: 870-883. doi:10.1111/j.1558-5646.2008.00602.x. PubMed: 19236476.
  57. 57. Moyle LC (2008) Ecological and evolutionary genomics in the wild tomatoes (Solanum sect. Lycopersicon). Evolution 62: 2995-3013. doi:10.1111/j.1558-5646.2008.00487.x. PubMed: 18752600.
  58. 58. Peralta I, Spooner D, Knapp S (2008) Taxonomy of wild tomatoes and their relatives (Solanum sect. Lycopersicoides, sect. Juglandifolia, sect. Lycopersicon; Solanaceae). Syst Bot Monogr 84: 1–186..
  59. 59. Städler T, Roselius K, Stephan W (2005) Genealogical footprints of speciation processes in wild tomatoes: demography and evidence for historical gene flow. Evolution 59: 1268-1279. doi:10.1554/04-722. PubMed: 16050103.
  60. 60. Spooner D, Peralta I, Knapp S (2005) Comparison of AFLPs with other markers for phylogenetic inference in wild tomatoes [Solanum. L. section Lycopersicon. (Mill.) Wettst.]. Taxon 54: 43-61.
  61. 61. Nakazato T, Warren DL, Moyle LC (2010) Ecological and geographic modes of species divergence in wild tomatoes. Am J Bot 97: 680-693. doi:10.3732/ajb.0900216. PubMed: 21622430.
  62. 62. Chetelat RT, Pertuzé RA, Faúndez L, Graham EB, Jones CM (2008) Distribution, ecology and reproductive biology of wild tomatoes and related nightshades from the Atacama Desert region of northern Chile. Euphytica 167: 77-93.
  63. 63. Lange BW, Langley CH, Stephan W (1990) Molecular evolution of Drosophila metallothionein genes. Genetics 126: 921-932. PubMed: 1981765.
  64. 64. Baudry E, Kerdelhué C, Innan H, Stephan W (2001) Species and recombination effects on DNA variability in the tomato genus. Genetics 158: 1725-1735. PubMed: 11514458.
  65. 65. Roselius K, Stephan W, Städler T (2005) The relationship of nucleotide polymorphism, recombination rate and selection in wild tomato species. Genetics 171: 753-763. doi:10.1534/genetics.105.043877. PubMed: 16085701.
  66. 66. Arunyawat U, Stephan W, Städler T (2007) Using multilocus sequence data to assess population structure, natural selection, and linkage disequilibrium in wild tomatoes. Mol Biol Evol 24: 2310-2322. doi:10.1093/molbev/msm162. PubMed: 17675653.
  67. 67. Städler T, Arunyawat U, Stephan W (2008) Population genetics of speciation in two closely related wild tomatoes (Solanum section Lycopersicon). Genetics 178: 339-350. doi:10.1534/genetics.107.081810. PubMed: 18202377.
  68. 68. Librado P, Rozas J (2009) DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451-1452. doi:10.1093/bioinformatics/btp187. PubMed: 19346325.
  69. 69. Watterson GA (1975) Number of segregating sites in genetic models without recombination. Theor Popul Biol 7: 256-276. doi:10.1016/0040-5809(75)90020-9. PubMed: 1145509.
  70. 70. Tajima F (1983) Evolutionary relationship of DNA sequences in finite populations. Genetics 105: 437-460. PubMed: 6628982.
  71. 71. Tajima F (1989) Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585-595. PubMed: 2513255.
  72. 72. Fu YX, Li WH (1993) Statistical tests of neutrality of mutations. Genetics 133: 693-709. PubMed: 8454210.
  73. 73. Fay JC, Wu CI (2000) Hitchhiking under positive Darwinian selection. Genetics 155: 1405-1413. PubMed: 10880498.
  74. 74. Depaulis F, Veuille M (1998) Neutrality tests based on the distribution of haplotypes under an infinite-site model. Mol Biol Evol 15: 1788-1790. doi:10.1093/oxfordjournals.molbev.a025905. PubMed: 9917213.
  75. 75. Lescot M, Déhais P, Thijs G, Marchal K, Moreau Y et al. (2002) PlantCARE, a database of plant cis-acting regulatory elements and a portal to tools for in silico analysis of promoter sequences. Nucleic Acids Res 30: 325-327. doi:10.1093/nar/30.1.325. PubMed: 11752327.
  76. 76. Expósito-Rodríguez M, Borges AA, Borges-Pérez A, Pérez JA (2008) Selection of internal control genes for quantitative real-time RT-PCR studies during tomato development process. BMC Plant Biol 8: 131. doi:10.1186/1471-2229-8-131. PubMed: 19102748.
  77. 77. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method. Methods 25: 402-408. doi:10.1006/meth.2001.1262. PubMed: 11846609.
  78. 78. Hellemans J, Mortier G, De Paepe A, Speleman F, Vandesompele J (2007) qBase relative quantification framework and software for management and automated analysis of real-time quantitative PCR data. Genome Biol 8: R19. doi:10.1186/gb-2007-8-2-r19. PubMed: 17291332.
  79. 79. R Development Core Team (2012) R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing.
  80. 80. Kirby DA, Stephan W (1996) Multi-locus selection and the structure of the white gene of Drosophila melanogaster. Genetics 144: 635-645. PubMed: 8889526.
  81. 81. Stern DL, Orgogozo V (2009) Is genetic evolution predictable? Science 323: 746-751. doi:10.1126/science.1158997. PubMed: 19197055.
  82. 82. Emerson JJ, Li W-H (2010) The genetic basis of evolutionary change in gene expression levels. Philos Trans R Soc B 365: 2581-2590. doi:10.1098/rstb.2010.0005. PubMed: 20643748.
  83. 83. Proost S, Van Bel M, Sterck L, Billiau K, Van Parys T et al. (2009) PLAZA: A comparative genomics resource to study gene and genome evolution in plants. Plant Cell 21: 3718-3731. doi:10.1105/tpc.109.071506. PubMed: 20040540.
  84. 84. Silhavy D, Hutvágner G, Barta E, Bánfalvi Z (1995) Isolation and characterization of a water-stress-inducible cDNA clone from Solanum chacoense. Plant Mol Biol 27: 587-595. doi:10.1007/BF00019324. PubMed: 7894021.
  85. 85. Padmanabhan V, Dias DAL, Newton RJ (1997) Expression analysis of a gene family in loblolly pine (Pinus taeda L.) induced by water deficit stress. Plant Mol Biol 35: 801-807. doi:10.1023/A:1005897921567. PubMed: 9426600.
  86. 86. Wang C-S, Liau Y-E, Huang J-C, Wu T-D, Su C-C et al. (1998) Characterization of a desiccation-related protein in lily pollen during development and stress. Plant Cell Physiol 39: 1307-1314. doi:10.1093/oxfordjournals.pcp.a029335. PubMed: 10050314.
  87. 87. Shen G, Pang Y, Wu W, Deng Z, Liu X et al. (2005) Molecular cloning, characterization and expression of a novel Asr gene from Ginkgo biloba. Plant Physiol Biochem 43: 836-843. doi:10.1016/j.plaphy.2005.06.010. PubMed: 16289880.
  88. 88. Philippe R, Courtois B, McNally KL, Mournet P, El-Malki R et al. (2010) Structure, allelic diversity and selection of Asr genes, candidate for drought tolerance, in Oryza sativa L. and wild relatives. Theor Appl Genet 121: 769-787. doi:10.1007/s00122-010-1348-z. PubMed: 20454772.
  89. 89. Schneider A, Salamini F, Gebhardt C (1997) Expression patterns and promoter activity of the cold-regulated gene ci21A of potato. Plant Physiol 113: 335-345. doi:10.1104/pp.113.2.335. PubMed: 9046587.
  90. 90. Blake WJ, Balázsi G, Kohanski MA, Isaacs FJ, Murphy KF et al. (2006) Phenotypic consequences of promoter-mediated transcriptional noise. Mol Cell 24: 853-865. doi:10.1016/j.molcel.2006.11.003. PubMed: 17189188.
  91. 91. Maheshri N, O'Shea EK (2007) Living with noisy genes: how cells function reliably with inherent variability in gene expression. Annu Rev Biophys Biomol Struct 36: 413-434. doi:10.1146/annurev.biophys.36.040306.132705. PubMed: 17477840.
  92. 92. Ha M, Li WH, Chen ZJ (2007) External factors accelerate expression divergence between duplicate genes. Trends Genet 23: 162-166. doi:10.1016/j.tig.2007.02.005. PubMed: 17320239.
  93. 93. Carey LB, van Dijk D, Sloot PM, Kaandorp JA, Segal E (2013) Promoter sequence determines the relationship between expression level and noise. PLOS Biol 11: e1001528. PubMed: 23565060.
  94. 94. the Tomato Genome Consortium (2012) The tomato genome sequence provides insights into fleshy fruit evolution. Nature 485: 635-641. doi:10.1038/nature11119. PubMed: 22660326.
  95. 95. He F, Zhang X, Hu J, Turck F, Dong X et al. (2012) Genome-wide analysis of cis-regulatory divergence between species in the Arabidopsis genus. Mol Biol Evol 29: 3385-3395. doi:10.1093/molbev/mss146. PubMed: 22641789.
  96. 96. Tellier A, Laurent SJY, Lainer H, Pavlidis P, Stephan W (2012) Inference of seed bank parameters in two wild tomato species using ecological and genetic data. Proc Natl Acad Sci U_S_A 108: 17052-17057. PubMed: 21949404.