1 Background
Olive (Olea europaea L.) is one of the most important cultivated tree crop species in Mediterranean regions (Rugini and Baldoni 2004). Tunisia is the most important olive-growing country of the southern Mediterranean region. Tunisian olive resources are estimated over 70 million olive trees, grown on 1 680 000 ha. Olives are found in all the regions of Tunisia, from North to South. Close to 90% of cultivated olive acreage is located in the central and southern regions. Tunisian areas are arid and semi-arid regions which indicate remarkable differences in the rates of precipitation which cause water stress. This region is characterized by a high rate of evapo-transpiration, high solar radiation and an irregular rainfall that does not surpass 200 mm per year. Periods with water restrictions as a result of climatic changes are frequent and problems of drought (due to non-irrigation) can appear. As olive cultivars may exhibit a different level of drought tolerance, the knowledge and the selection of the most drought-tolerant cultivars acquires relevance. The molecular basis of dehydration tolerance in plant is extremely complex and a wide variety of expressional candidate genes and metabolic pathways has been suggested during drought stress (Umezawa et al. 2006; Syed Sarfraz et al. 2011). Consequently, much effort has been directed towards a better understanding of the genetic basis of the adaptive response of plants to drought, and how best to apply this knowledge to molecular plant breeding. The physiological mechanisms and the genetic basis of drought stress tolerance in olive tree are scarce. For these reasons, advancement in the current understanding of the responses of olive trees undergo severe stress through seasonal drought has become a major target for research in Tunisia. In this respect, the selection of cultivars with enhanced water use efficiencies and drought tolerance is one of the most effective strategies for improving productivity and for their survival in drought field conditions. To this end, the development of a set of suitable informative SNP DNA markers is important, in order to provide a technique for rapidly screening cultivars with different drought tolerances.

Aquaporins (AQPs) are known to facilitate the movement of water and small solutes across cellular membranes. MIP (major intrinsic protein) like isoforms genes involved in drought tolerance mechanisms have been identified, characterized, and assessed for their comparative transcriptional activity by using whole- genome sequencing or expressed sequence tag (EST) libraries in different model species such as Arabidopsis (Johanson et al. 2001) and maize (Chaumont et al. 2001). Higher plant aquaporins proteins represent a large family of the major intrinsic protein (MIP) superfamily. Aquaporins consist of five subfamilies that include; 1- plasma membrane intrinsic proteins (PIP), 2- tonoplast intrinsic proteins (TIP), 3- NOD26-like intrinsic proteins (NIP), 4- small basic intrinsic proteins (SIP) and 5- the recently identified X (or unrecognized) intrinsic proteins (XIP) (Park et al. 2010).

Aquaporins are involved in the regulation of water flow and have been shown to be involved in drought response. Recently, Secchi et al. (2007a,b) deeply investigated the molecular mechanisms and the identification transcription factors, including genes related to stress response and water transport in olive. They concluded that aquaporins play a major adaptive role in olive. Indeed, they analyzed the change in the expression level of genes related to the aquaporin family in olive subjected to drought treatment. Three aquaporin (AQPs) genes have been isolated from cv Leccino tissues exposed to different environmental conditions. The authors found a strong down regulation in these genes following drought stress, probably resulting in reduced membrane water permeability and preventing the loss of water in periods of water stress (Bracci et al. 2011).

Moreover, many genes were identified and annotated based on EST sequencing in some olive cultivars, which are involved on fatty acid biosynthesis (Haralampidis et al. 1998; Hatzopoulos et al. 2002), triacylglycerols (TAGs) biosynthesis (Banilas et al. 2010), antioxidant biosynthesis (Saimaru et al. 2007), fruit growth and ripening processes (Martinelli and Tonutti 2012), pollen allergens (Hamman-Khalifa et al. 2008) and drought stress response (Secchi et al. 2007a,b).

In this study, we describe the genetic diversity in the TIP and PIP2 sub-family of the widespread gene family of aquaporins in 15 commercial Tunisian olive cultivars.

2 Results and discussion
2.1 SNPs frequency and nucleotide diversity
Aquaporins candidate genes (TIP and PIP2) were selected from literature and in silico analysis attending to their putative role in drought stress response. The 2 genomic loci were amplified and were properly sequenced in 15 genotypes representative of cultivated olive.

germplasm from distant regions in the country. Different patterns of polymorphisms were obtained. All discovered sequences for the two used markers reported in FASTQ are available in the ENA data- base (http://www.ebi.ac.uk/ena) under the accessions number: for TIP gene (Chamlali: HG965187, Zarrazi: HG965188, Zalmati: HG965189, Chétoui: HG965190, Toffehi: HG965191, Chamlali-Jerba: HG965192 and Neb-Jamel HG965193) and for PIP2 gene: (Chamlali-Jerba: HG965196). Summarizing, the levels of DNA polymorphism were found in Table 2. SNP discovery was validated by genotyping a set of 15 cultivars using direct sequencing of PCR products and comparative analysis of sequences. A total of 6 candidate regions related to drought stress were selected for SNP identification and nucleotide diversity analyses. These two loci comprise 1583 bp of aligned sequence per individual. TIP and PIP2 genes ranged from 519 and 1094 pb respectively. In total, we found 119 SNP in both coding and non-coding regions. The two loci used in this study were polymorphic and the SNP frequency detected in our set of olive cultivars was 1 SNP/13.30 bp. Thus, an average of 1 SNP every 9.8 bp for TIP gene and 1 SNP every 16.8 pb for PIP2 for all regions were detected. The level of SNPs discovered in this study was higher than previous reports on olive genetic DNA analysis (5 SNPs in Santos Macedo et al. (2009); 9 SNPs in hakim et al. (2010)). These results are in contrast comparable to the frequency obtained by Besnard and El Bakkali (2014) who found on two olive subspecies one substitution per 100 bp. On the other hand, other agronomically important crops like sunflower (1 SNP/69 bp) (Fusari et al. 2008) and cultivated and wild grapevine (1 SNP/63) (Riahi et al. 2013) presented a higher SNP frequency than the olive cultivars surveyed in this work. Nevertheless, the discrepancy between nucleotide genetic studies could be caused by differences in gene sampling. Moreover, the number of SNPs varied also between coding and non-coding regions: 53 SNPs were found in 392 bp of coding regions whereas 66 SNPs were detected in 1191 bp of non-coding sequences: hence, the SNP frequency was 1 SNP/7.39 bp in coding regions and 1SNP/18.04 bp in non-coding regions. These results suggest that coding regions are more conserved (less SNP frequency) than non-coding regions, most probably due to purifying selection (Fusari et al. 2008). Genetic variation at the nucleotide level was estimated from nucleotide diversity (π = 0.0024 and π = 0.0017) and from the number of segregating sites as indicated by Watterson (1975) (θW = 0.0020 and θW = 0.0013) for TIP and PIP2 genes respectively. The recorded level of nucleotide diversity in this study is comparable to that reported by Audigeos et al. (2010) in Neotropical trees belonging to Eperua falcata and Virola sebifera species which the genetic diversity ranged between 2.20 and 2.72 and higher than in Eperua grandiflora (π=1.21) in aquaporins PIP gene family studies.

The average level of nucleotide diversity was lower than those observed in two wild olive subspecies (suspp. Europaea and cuspidata), (π=0.0057; π=0.0031, respectively) (Besnard and El Bakkali 2014). Indeed, Sabetta et al. (2013) showed higher nucleotide diversity at nuclear genes as ω-3 fatty acid desaturases gene (Fad7) in Italian olive cultivars. Nevertheless, the discrepancy between olive studies on nucleotide diversity could be explained by the genetic divergence of the materials analyzed as olea species have a wide genetic diversity and could be caused by the difference of genomic sequenced regions since genomic analysis, generally depends on the nature of the study and their fittingness to sample different parts of the genome. Indeed, in this study we used generated primers from highly conserved regions of candidate genes involved in drought response from model species. The transfer ability of information for olive could be a source of variation due the scarce of olive genome sequences, the big size of olive genome (1,800 Mb), is a diploid species (2n = 2x = 46) and predominantly allogamous (Rugini et al. 2011). This difference is not surprising as studied cultivars were selected for covering varying geographic and environmental ranges and as observed by Secchi et al. (2007a), there is a relationship between the expression patterns of AQPs and physiological responses in olive during drought and up and down-regulation of this protein allows plants to respond to environmental changes and to maintain their water status which intrinsic plant factors depend from development of plant architecture and rootstock (Secchi et al. 2007b).

The average nucleotide polymorphism and nucleotide diversity of non-coding regions (π = 0.0020) was slightly lower, although non-significant, than diversity estimates in coding regions (π = 0.0021). Average synonymous-sites (π = 0.0026) was lower than non-synonymous (π = 0.0028) for the two genes suggesting that the diversity of these regions is governed by purifying selection and an excess of synonymous mutations was occurred in this candidate gene.

A trend for a higher observed homozygosity (65%) and expected heterozygosity (24.3%) in TIP gene was observed than in PIP2 gene (Ho = 55%; He= 18.7% respectively). As indicated by Riahi et al. (2013), the SNP molecular marker system show usually lower expected heterozygosity due to its bi-allelic nature. The genetic variability was higher in TIP gene than in PIP2 locus for the 15 Tunisian cultivars determined by SNP molecular markers. The high levels of diversity at TIP gene than PIP2 gene reveal more details on the processes shaping olive diversity and provide information on the link between genetic diversity and ecological conditions. However, the non uniformity between the length regions sequences (coding, non-coding) for the two studies genes could influence different molecular processes in plant and statistical tests seem to be mostly restricted at the exon-intron boundary (Audigeos et al. (2010).

2.2 Neutrality tests
Neutrality tests were applied to identify departures from the standard neutral patterns of evolution in the candidate loci, like Tajima’s D (1989), Fu and Li’s tests (1993) and Fu’s Fs (1997). Tajima’s D‐statistic was computed for each locus and reflects the difference between π and θW. Fu and Li’s and Fu’s Fs neutrality tests were computed without outgroup.

For TIP and PIP2 genes, allele frequency distribution is indicated by the value of Tajima's D which was positive, not significantly and different from 0. As showed by Audigeos et al. (2010) positive Tajima's D value indicates a deficit of low frequency alleles relative to neutral expectations in a randomly mating population of constant size and could be the consequence of population bottlenecks, population subdivision or balancing selection as would be expected in breeding populations. Fu and Li’s tests and Fu’s Fs neutrality tests were positive for the two genes suggesting a deficiency of recent mutations and a lower number of haplotypes than expected. Fu’s F test was negative for PIP2 gene (-0.129) indicating that the departure from neutrality is only apparent. Interpretation of the neutrality tests is difficult since many recombination events could occur in natural populations. Fu (1996) note that recombination events reduce the variances of nucleotide genetic diversity estimators and increase the number of alleles in a sample. Accordingly, further research on the genetic diversity of Tunisian olive by SNP analysis related to drought stress is needed to investigate other Tunisian hybrids and rootstocks. The attainment of better levels of discrimination of olive cultivars related to their environment requires the testing of more candidate genes and more primers via classic sequencing methods as well as using NGS platform.

2.3 Genetic relationships
In order to analyze the genetic relationships among studied cultivars for the two genes, dendrograms were constructed based on the UPGMA cluster analysis of SNP data. The clustering patterns obtained are given in Figures 1–2.

Figure 1 UPGMA dendrogram showing the discrimination of the studied cultivars based on TIP DNA sequence

Figure 2 UPGMA dendrogram showing the discrimination of the studied cultivars based on PIP2 DNA sequence.

The analysis of the SNP data for TIP gene revealed a clear separation of cultivars into 3 groups by cutting the dendrogram at a genetic similarity value of 0.17, while the dendrogram in PIP2 gene separate cultivars in two clear separate groups, with greatest separation of cultivars. Our results failed to differentiate cultivars into separate clusters according to their susceptibility to drought (resistant vs. susceptible). No further discrimination was, however, visible among the different varieties to their geographical origins (south, north or center of Tunisia) and environmental ranges for the two genes. However, an overlap between groups is reported. Thus, we note the clustering of some cultivars from the north in group composed by southern varieties was showed in the two data. Moreover, it could be found that cultivars from different regions and different susceptibility to water scarcity such as ‘Zarrazi Zarzis’, ‘Fokhari’, ‘Chamlali Sfax’, ‘Chamlali-Jerba’, ‘Zalmati’ and ‘Toffehi’ which are  “resistant” and ‘Chétoui’ and ‘Neb-Jamal’ are “sensitive” were clustered together. These results show a high genetic variation and high degree of polymorphism in Tunisian olive cultivars. Comparable results were found by Ercisli et al. (2011), who shown that Turkish olive cultivars exhibit a high genetic diversity by using SSR molecular markers techniques. Meanwhile, some studied cultivars originated from the southern of Tunisia considered as “medium resistant” cultivars were clustered with cultivars from the north. Probably, amplified TIP and PIP2 loci were not closely associated with major loci controlling drought resistance in olive. These results were in agreement with catalase-based markers applied in apple genetic analysis and PCR-based strategy failed to discriminate resistant and susceptible apple accessions to scab and powdery mildew responses (Gulsen et al. 2010).

The results presented here were the first report on molecular diversity in aquaporines genes of commercially relevant species of one of the most diverse in Mediterranean countries. AQPs-based markers may be applied in Tunisian olive germplasm genetic diversity analysis. The fragments of TIP and PIP2 genes analyzed here have revealed moderate large variability with patterns that vary along the gene. Our results demonstrated that genetic variation at studied aquaporin genes may represent a response to variable environmental conditions. Knowledge about genetic relationships between Tunisian olive cultivars could be helpful in olive improvement strategies and could facilitate association mapping in olive. An extended characterization of these loci and other loci involved in drought stress could provide more information on the link among genetic diversity, environmental conditions and adaptive traits related to water stress.

3 Material and methods
3.1 Plant material and DNA extraction
The study was conducted on 15 main commercial Tunisian olive cultivars which are listed in Table 1. Most of the plant materials were obtained from the National Olive Genetic Resources Conservation located in Boughrara-Sfax. Cultivars were selected for covering varying geographic and environmental ranges with different amount of precipitation in Tunisia (Table 1).

Table 1 List of Tunisian olive cultivars included in this study

Total genomic DNA was extracted from small leaves using hexadecyl trimethyl ammonium bromide (CTAB) according to the method described by Doyle and Doyle (1990).

DNA was purified, quantified, suspended in a TE solution (pH 8) and stored at –20℃.

3.2 Molecular analysis
Two genes namely tonoplast intrinsic protein (TIP, accession N°: DQ202710.1) and plasma membrane intrinsic protein (PIP2, accession N°: DQ202708.1) from aquaporin gene family (AQPs) which are involved in olive drought response (Secchi et al. 2007b) were selected for SNPs characterization on 15 Tunisian olive varieties (Table 2). Reference DNA sequences of these genes were obtained from the National Centre for Biotechnology Information (NCBI) GenBank database (www.ncbi.nlm.nih.gov).

5 primers for TIP and 6 primers for PIP2 genes were designed using the program Primer3 (v.0.4.0; Rozen and Skaletsky 2000) in conserved regions. Only one primer for each gene was selected for his usefulness in genotyping olive cultivars. The forward and reverse primer for each gene was indicated in Table 2.

Table 2 Genetic diversity parameters and neutrality tests for the two studied genes

PCR amplification was done in total volume of 25 μL containing 1x PCR reaction buffer, 10 μM of each primer, 0.20 mM dNTPs, and 3U of DNA Taq polymerase. The PCR was carried out on a BIORAD programmable thermal controller as follows: initial denaturation at 95 °C for 2 min; 35 cycles of 94 C° for 30 sec, an annealing step at 56-61C° for 30 sec (For each candidate gene, a primer optimization step was done on three genotypes) and 72 C° for 1.30 min, and final extension at 72 C° for 5 min.

The expected size of each PCR product was confirmed by separation on 1.5% agarose gel, stained with ethidium bromide, and visualized under UV light. The amplified PCR products were purified and sequenced on an ABI sequencing instrument.

3.3 Data analysis
DNA sequences were examined visually as well as for the analysis of SNPs within the genes, the sequences were aligned using BioEdit version 7.0.9.0 (Hall 1999) using ClustalW mulTIPle alignment (Thompson et al. 1994).

Analyses of sequence data were performed using DnaSP v. 4.0 (Rozas et al. 2003). The nucleotide diversity (π) is based on the average number of nucleotide differences per site between sequences (Nei and Li 1979), whereas (θw) is based on the number of segregating sites (Watterson 1975) were calculated.

Tajima’s D (1989), Fu and Li’s (1993) and Fu’s Fs (1997) neutrality tests were conducted by using LDnaSP v. 4.0. Tajima’s D test was calculated for each locus and reflects the difference between π and θW. Fu and Li’s and Fu’s Fs tests were computed without outgroup.

Dendrograms were constructed based on the distance matrix with the UPGMA method using the MEGA program (Version 4, Tamura et al. 2007). Bootstrap (1000 replications) analysis was performed to estimate the confidence of the topology of the consensus tree.

4 Acknowledgements
This project was funded by the Technical Cooperation Project on Korea-Africa Food and Agricultural Cooperation Initiative between the Rural Development Administration of the Republic of Korea and the Republic of Tunisia. We are grateful to Dr I.S. Yoon and Dr M.O. Byun for providing technical help during the research.

References
Audigeos D., Buonamici A., belkadi L., Rymer P., Boshier D., Scotti-Saintagne C., Vendramin G.G., and Scotti I., 2010, Aquaporins in the wild: Natural genetic diversity and selective pressure in the PIP gene family in five Neotropical tree species, BMC Evolutionary Biology, 10: 202
http://dx.doi.org/10.1186/1471-2148-10-202
  
Banilas G., Karampelias M., Makariti I., Kourti A., and Hatzopoulos P., 2010, The olive DGAT2 gene is developmentally regulated and shares overlapping but distinct expression patterns with DGAT1, Journal of Experimental Botany, 62: 521-532
http://dx.doi.org/10.1093/jxb/erq286

Besnard G., and El Bakkali A., 2014, Sequence analysis of single-copy genes in two wild olive subspecies: nucleotide diversity and potential use for testing admixture. Genome, 57: 145–153
http://dx.doi.org/10.1139/gen-2014-0001

Bracci T., Busconi M., Fogher C., and Sebastiani L., 2011, Molecular studies in olive (Olea europaea L.): overview on DNA markers applications and recent advances in genome analysis, Plant Cell Reports, 30: 449–462
http://dx.doi.org/10.1007/s00299-010-0991-9

Chaumont F., Barrieu F., Wojcik E., Chrispeels M.J., and Jung R., 2001, Aquaporins constitute a large and highly divergent protein family in maize, Plant Physiology, 125: 1206–1215
http://dx.doi.org/10.1104/pp.125.3.1206

Doyle J.J., and Doyle J.L., 1990, Isolation of plant DNA from fresh tissue, Focus, 12: 13–15

Ercisli S., Ipek A., and Barut E., 2011, SSR Marker-Based DNA Fingerprinting and Cultivar Identification of Olives (Olea europaea), Biochemical Genetics, 49: 555–561
http://dx.doi.org/10.1007/s10528-011-9430-z

Fu Y.X., 1996, New Statistical Tests of Neutrality for DNA Samples from a Population, Genetics, 143: 557‐570

Fu Y.X., 1997, Statistical Tests of Neutrality of Mutations against Population Growth, Hitchhiking and Background Selection, Genetics, 147: 915‐925

Fu Y.X., and Li W.H., 1993, Statistical Tests of Neutrality of Mutations, Genetics, 133: 693‐709

Fusari C.M., Lia V.V., Hopp H.E., Heinz R.A., and Norma B Paniego N.B., 2008, Identification of Single Nucleotide Polymorphisms and analysis of Linkage Disequilibrium in sunflower elite inbred lines using the candidate gene approach, BMC Plant Biology, 8: 7. doi:10.1186/1471-2229-8-7
http://dx.doi.org/10.1186/1471-2229-8-7

Gulsen O., S Kaymak S., Ozongun S., and Uzun A., 2010, Genetic analysis of Turkish apple germplasm using peroxidase gene-based markers, Scientia horticulturae, 125: 368–373
http://dx.doi.org/10.1016/j.scienta.2010.04.023

Hakim R.I., Grati-Kammoun N., Makhloufi E., and Rebaï A., 2010, Discovery and potential of SNP markers in characterization of Tunisian olive germplasm, Diversity, 2: 17–27
http://dx.doi.org/10.3390/d2010017

Hall T.A., 1999, BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT, Nucleic Acids Symposium Series, 41: 95–98

Hamman-Khalifa A., Castro A.J., Jiménez-Lόpez J.C., Rodríguez-García M.I., and De Dios Alché J., 2008, Olive cultivar origin is a major cause of polymorphism for Ole e 1 pollen allergen, BMC plant Biology, 8:10
http://dx.doi.org/10.1186/1471-2229-8-10

Haralampidis k., Milioni D., Sanchez J., Baltrusch M., Heinz E., and Hatzopoulos P., 1998, Temporal and transient expression of stearoyl-ACP carrier protein desaturase gene during olive fruit development, Journal of Experimental Botany, 327: 1661–1669

Hatzopoulos P., Banilas G., Giannoulia K., Gazis F., Nikoloudakis N., Milioni D., and Haralampidis K., 2002, Breeding, molecular markers and molecular biology of the olive tree, European Journal of Lipid Science and Technology, 104: 574–586
http://dx.doi.org/10.1002/1438-9312(200210)104:9/10%3C574::AID-EJLT574%3E3.0.CO;2-1

Johanson U., Karlsson M., Johansson I., Gustavsson S., Sjovall S., Fraysse L., Weig A.R., and Kjellbom P., 2001, The complete set of genes encoding major intrinsic proteins in Arabidopsis provides a framework for a new nomenclature for major intrinsic proteins in plants, Plant Physiology, 126: 1358–1369
http://dx.doi.org/10.1104/pp.126.4.1358

Martinelli F., and Tonutti P., 2012, Flavonoid metabolism and gene expression in developing olive (Olea europaea L.) fruit, Plant Biosystems, 146: 164-170
http://dx.doi.org/10.1080/11263504.2012.681320

Nei M., and Li W.H., 1979, Mathematical model for studying genetic variation in terms of restriction endonucleases, Proceedings of the National Academy of Sciences of the United States of America, 76: 5269-5273
http://dx.doi.org/10.1073/pnas.76.10.5269

Riahi L., Zoghlami N., Dereeper A., Laucou V., Mliki A., and This P., 2013, Single nucleotide polymorphism and haplotype diversity of the gene NAC4 in grapevine, Industrial Crops and Products, 43: 718–724
http://dx.doi.org/10.1016/j.indcrop.2012.08.021

Park W., Scheffler B.E., Bauer P.J., and Campbell B.T., 2010, Identification of the family of aquaporin genes and their expression in upland cotton (Gossypium hirsutum L.), BMC Plant Biology, 10: 142 doi:10.1186/1471-2229-10-142
http://dx.doi.org/10.1186/1471-2229-10-142

Rosen S., and Skaletsky H., 2000, Primer3 on the WWW for general users and for biologist programmers, Methods in Molecular Biology, 132: 365–386

Rozas J., Sanchez-Delbarrio J.C., Messeguer X., and Rozas R., 2003, DnaSP DNA polymorphisme analyses by the coalescent and other methods, Bioinformatics ,19: 2496–2497
http://dx.doi.org/10.1093/bioinformatics/btg359

Rugini E., and Baldoni L., 2004, Olea europaea olive. In: Biotechnology of Fruit and Nut Crops. (Litz, RE, Ed.), CABI Publishing,Wallingford, Oxon., UK, 404–428

Rugini E., De pace C., Gutiérrez-Pesce P., and Rosario M., 2011, wild crop relatives: Genomic and breeding resources Temperate fruits: Olea Springer, New York, P241

Sabetta W., Blanco A., Zelasco S., Lombardo L., Perri E., Mangini G., and Montemurro C., 2013,  Fad7 gene identification and fatty acids phenotypic variation in an olive collection by EcoTILLING and sequencing approaches, Plant Physiology and Biochemistry, 69:1–8
http://dx.doi.org/10.1016/j.plaphy.2013.04.007

Saimaru H., Orihara Y., Tansakul P., Kang Y.H., Shibuya M., and Ebizuka Y., 2007, Production of triterpene acids by cell suspension cultures of Olea europaea, Chemical & pharmaceutical bulletin, 55: 784–788
http://dx.doi.org/10.1248/cpb.55.784

Santos Macedo E., Cardoso H.G., Hernández A., Peixe A.A., Polidoros A., Ferreira A., Cordeiro A.,
and Arnholdt-Schmitt B., 2009, Physiologic responses and gene diversity indicate olive alternative oxidase as a potential source for markers involved in efficient adventitious root induction,  Physiologia Plantarum, 137: 532–552
http://dx.doi.org/10.1111/j.1399-3054.2009.01302.x

Secchi F., Lovisolo C., Uehlein N., Kaldenhoff  R., and Schubert A., 2007a, Isolation and functional characterization of three aquaporins from olive (Olea europea L), Planta, 225: 381-392
http://dx.doi.org/10.1007/s00425-006-0365-2

Secchi F., Lovisolo C., and Schubert A., 2007b, Expression of OePIP2.1 aquaporin gene and water relations of Olea europaea twigs during drought stress and recovery, Annals of Applied Biology, 150: 163–167
http://dx.doi.org/10.1111/j.1744-7348.2007.00118.x

Syed Sarfraz H., Akhtar Kayani M., and Amjad M., 2011, Transcription factors as tools to engineer
enhanced drought stress tolerance in plants, Biotechnology Progress, 27: 297-306
http://dx.doi.org/10.1002/btpr.514

Tajima F., 1989, Statistical Method for Testing the Neutral Mutation Hypothesis by DNA Polymorphism, Genetics, 123: 585‐595

Tamura K., Dudley J., Nei M., and Kumar S., 2007, MEGA4: Molecular Evolutionary Genetics
Analysis (MEGA) software version 4.0, Molecular Biology and Evolution, 24: 1596–1599
http://dx.doi.org/10.1093/molbev/msm092

Thompson J.D., Higgins D.G., and Gibso T.J., 1994, CLUSTAL W: Improving the sensitivity of progressive mulTIPle sequence alignment through sequence weighting position-specific gap penalties and weight matrix choice, Nucleic Acids Research, 22: 4673–4680
http://dx.doi.org/10.1093/nar/22.22.4673

Umezawa T., Fujita M., Fujita Y., Yamaguchi-Shinozaki K., and Shinozaki K., 2006, Engineering
drought tolerance in plants: discovering and tailoring genes to unlock the future, Current Opinion in Biotechnology, 17: 113–122
http://dx.doi.org/10.1016/j.copbio.2006.02.002
 
Watterson  G.A., 1975, On the number of segregating sites in genetical models without recombination,
Theoretical Population Biology, 7: 188–193
http://dx.doi.org/10.1016/0040-5809(75)90020-9