Genetics of Na⁺ exclusion and salinity tolerance in Afghani durum wheat landraces

Plants vary widely in tolerance to salt stress, with durum wheat being particularly sensitive [1]. Commercial durum wheat is grown in temperate regions of the world, under both irrigated and rain-fed systems [2]. Much of the world’s arable land, however, is affected by salinity, and this is a major limitation for durum wheat production [3, 4]. Low shoot Na⁺ concentration is necessary for normal growth of most cereals. Indeed, maintenance of a high K⁺/Na⁺ ratio in the shoot was identified as one of the major mechanisms associated with salinity tolerance in the Triticeae and other plants [5]. Tetraploid durum wheat cultivars (Triticum turgidum ssp. durum) are more sensitive to salinity than hexaploid bread wheat (T. aestivum) and this is associated with their high concentration of Na⁺ in the shoots [3, 4].

Several attempts have been made to improve Na⁺ exclusion and the ratio of K⁺ / Na⁺ in durum wheat. Natural variability for salinity tolerance among durum wheat cultivars is very limited [6] and this can be increased by introducing traits from the D-genome of bread wheat or through the use of landraces or wild species [7]. The Kna1 locus on the long arm of chromosome 4D in bread wheat, found to be responsible for maintaining beneficial K⁺/Na⁺ ratios, was introgressed into the long arm of chromosome 4B in durum wheat using disomic substitution lines [7]. Although a novel durum line with lower shoot Na⁺ concentration and a higher shoot K⁺/Na⁺ ratio was generated, this line produced a lower biomass compared with other durum wheat cultivars. It was later found that Kna1 was also linked with undesirable traits [8].

Another approach to enhance Na⁺ exclusion in durum wheat involves the screening of large numbers of accessions and landraces to identify new sources of Na⁺ exclusion which can be introduced into elite durum cultivars. Wild relatives of wheat have been used for many years in breeding programmes as a source for novel alleles and traits for improving yield under abiotic and biotic stress. In the area of salinity research, the Indian landrace Kharchia 65 was found to have high exclusion of Na⁺ from the leaves [9] and has been widely used to develop salt tolerant breeding lines and commercial cultivars of bread wheat [10]. Bread wheat landraces from Iran, Nepal and Pakistan have been found to contain novel sources of salinity tolerance which have been used to improve yield in saline fields [11, 12].

However, durum wheat landraces have only rarely been used for the improvement of salinity tolerance. A source of Na⁺ exclusion was identified in durum wheat line 149, which was derived from a cross between a wild diploid species T. monococcum, accession C68–101, and the durum wheat cv. Marrocos [6]. The shoot Na⁺ exclusion phenotype was associated with two loci, one on the long arm of chromosome 2A (Nax1) [13] and the second on the long arm of chromosome 5A (Nax2) [14]. These loci were the site of an old translocation from T. monococcum [15]. Further investigation revealed that the Nax1 locus is associated with Na⁺ exclusion by unloading Na⁺ in the leaf sheath so reducing leaf blade Na⁺ and increasing leaf blade K⁺. This locus is also associated with the retrieval of Na⁺ ions from the xylem in the roots before they can be transported to the shoots. The second locus, Nax2, is also associated with the retrieval of Na⁺ from the xylem, although only in the roots, before the ions can be transported to the shoots [14]. The candidate genes for the Nax1 and Nax2 loci were identified as TmHKT1;4-A and TmHKT1:5-A, respectively. TmHKT1;5-A, is a homoeolog to the bread wheat gene, TaHKT1;5-D, which is the candidate gene for the Kna1 locus [15, 16]. The introgression of TmHKT1;5 (Nax2) into near-isogenic tetraploid lines with a cv. Tamaroi background resulted in a yield enhancement of 26% under salinized field conditions [17].

In earlier work, two Afghani durum landraces, Gandum Siahloshe Zamistani Aubi (AUS-14740) and Gandum Kofari (AUS-14752), were found to accumulate approximately half the concentration of Na⁺ in the third leaf compared with the other 177 durum wheat cultivars and landraces screened [18]. Despite their common geographical origin, lines AUS-14740 and AUS-14752 had different growth habits and plant height, corresponding presumably to different genetic backgrounds. Therefore, it is possible that lines AUS-14740 and AUS-14752 may be sources of novel alleles of genes controlling Na⁺ exclusion and a high K⁺/Na⁺ ratio. The aim of the present study was to determine the genetic control of Na⁺ exclusion and the K⁺/Na⁺ ratio in segregating populations of durum wheat derived from these two exotic Afghani landraces.

Genetic loci for Na⁺ exclusion and K⁺/Na⁺ ratio were identified in a durum mapping population, produced by a cross between the elite Australian cultivar, Jandaroi and an Afghani exotic durum landrace, AUS-14740, which accumulates in leaves half the Na⁺ concentration compared to other durum wheats.

In the current study, the distributions of all traits (third leaf Na⁺ concentration, third leaf K⁺ concentration, and the K⁺/Na⁺ ratio) in both populations was suggestive of a bimodal indicating a possible monogenic segregation. These findings suggest that single major genes are associated with Na⁺ exclusion in both the Afghani durum landraces studied. Heritability for the traits studied was high in both populations, with the exception of third leaf K⁺ concentration. A moderate to high heritability for shoot Na⁺ concentration has previously been estimated in two F₂ populations of durum wheat [19]; two major QTL, Nax1 for HKT1;4 and Nax2 for HKT1;5, had been then identified in the same populations [13, 15]. Markers flanking these putative QTLs were used to develop salt tolerant durum wheat cultivars [17].

In order to define the genetic regions and identify potentially useful markers associated with third leaf Na⁺ concentration in the durum wheat of Afghani landraces, bulked segregant analysis (BSA) was conducted using SNP markers in the F₂ population originating from the cross Jandaroi × AUS-14740.

In the Jandaroi × AUS-14740 F₂ population, two SNP markers were found to be strongly associated with Na⁺ concentration in the leaves. The first marker, Xm5511, located on the long arm of chromosome 3B, was associated only with third leaf Na⁺ concentration and with neither third leaf K⁺ concentration nor the K⁺ /Na⁺ ratio. In contrast, a second marker, Xm564, was identified in the distal region of the long arm of chromosome 4B, and found to have a strong association with all three traits studied (Table 2).

Other QTL associated with these traits have been detected in other studies. The Kna1 locus was mapped to the distal region of chromosome 4DL in bread wheat [20] and this locus is associated with the selective accumulation of K⁺ over Na⁺ in the shoot. In durum wheat, the Nax2 locus was mapped to a region of the long arm of the chromosome 5A and this is associated with Na⁺ concentration and the ratio of K⁺/Na⁺ in the leaf blade [19]. It was found that Nax2 encoded a similar gene as Kna1 [15]. In rice (Oryza sativa), a QTL for shoot K⁺ content was mapped onto chromosome 1 in an F₂ population derived from a cross between a salt tolerant indica (Nona Bokra) and an intolerant elite Japonica variety (Koshihikari). This QTL was named SKC1 [21]. Low Na⁺ concentration in the leaves, improved selection of K⁺ over Na⁺ transported from roots to shoots, and a high ratio of K⁺/Na⁺ in the leaves are common phenotypic characteristics between Nax2, Kna1 and SKC1 [15]. As mentioned above, these characteristics were seen in the F₂ population derived from Jandaroi and AUS-14740. A candidate gene of the SKC1 locus is OsHKT1;5, which reduces the amount of Na⁺ transported from the root to the shoot [21]. An HKT-like gene was also cloned in Kna1 and Nax2, and HKT1;5 is a candidate for both loci; named as TaHKT1;5-D and TmHKT1;5-A, respectively [15]. Kna1 is found in bread wheat (T. aestivum L.) whereas Nax2 was derived from T. monococcum. It is possible that the Xm564 locus identified in this study corresponds to the Kna1 and Nax2 loci and the candidate gene for the locus on the long arm of chromosome 4B is an HKT-like gene. Huang, et al. [22] found four bands of HKT1;5-like genes in bread wheat using a HKT1;5 probe in Southern blots. Three hybridisation bands were located on the long arm of chromosome 4B and one band on the long arm of chromosome 4D. The HKT1;5 gene has only been found in bread wheat and T. monococcum, and has not been identified in durum wheat [17]. The results suggest that Na⁺ exclusion in AUS-14740 may be due to the gene HKT1;5.

The locations of both markers on chromosomes 3BL and 4BL identified in the current study and known HKT1;5 genes on chromosome 5AL in T. monococcum and on chromosome 4DL in bread wheat are shown in Fig. 2. Both homoeologous genes in the A and D genomes have similar genetic locations along the corresponding chromosomes. HKT1;5-A (Nax2) has been mapped to the distal part of chromosome 5A, with an estimated genetic distance of about 17 cM from the end of chromosome 5AL [15]. Part of chromosome 5AL originated from part of chromosome 4AL due to an ancient reciprocal translocation between the distal ends of chromosomes 4AL and 5AL [23]. HKT1;5-D (Kna1) has been identified in the genetic region about 14% from the end of chromosome 4DL [20].

We used the IWGSC website (http://pgsb.helmholtz-muenchen.de/plant/wheat//iwgsc/index.jsp) and the Gramene website to blast sequences from the Infinium design, where each SNP marker (used in our study) was developed, against the sequence of wheat chromosome 4B. We identified Xm564 markers on the distal part of chromosome 4BL in wheat contig 7,009,377, along with another marker, wsp_BG604404B. The position of Xm564 was 140.39 cM, out of a total length of 157.59 cM for the chromosome 4B genetic map, the marker being located about 17.1 cM from the end of chromosome 4B (Fig. 2). The HKT1;5-B1 like gene is also located on the wheat contig 7,009,377 and is in a very strong candidate for gene underlying this locus. This hypothesis, however, needs to be tested by looking for variation in or around the target gene by genotyping HKT1;5-B1 in the F₂ progeny from the cross Jandaroi × AUS-14740. In future, sequencing of the HKT1;5-B1 gene from AUS-14740 and Jandaroi would determine whether there were differences in the protein encoded by the two alleles. If there were differences, heterologous expression of the two alleles of HKT1;5-B1 in yeast and Xenopus oocytes would investigate whether differences in coding sequence resulted in altered function of the protein and its affinity for the Na⁺ ion. In addition, quantitative RT-PCR would determine whether the differences in the Na⁺ phenotype observed between AUS-14740 and Jandaroi is due to differences in spatial and/or temporal expression the HKT1;5-B1 gene between the two accessions. It would also be important to make a separate study of the Jandaroi × AUS-14752 segregating population, as the two exotic Afghani landraces may contain different alleles of HKT1;5-B1 gene.

Using genetic similarity between rice and wheat genes, we were able to localise the location of a second HKT1;5 gene on chromosome BL, HKT1;5-B2. This gene, however, is found in the near-centromeric region of the chromosome 4BL and is unlikely to be responsible for the phenotype linked to marker Xm564 (Fig. 2).

The role played by other possible genes remains unclear. SNP marker Xm5511 was located on the chromosome 3BL, and it was found to make a similar, high, contribution to Na⁺ concentration in the third leaf (Table 2). We mapped the marker Xm5511 to the distal part of chromosome 3BL, with a physical location of about 701.9 Mbp, out of the 774.4 Mbp total length of the chromosome 3B physical map (Fig. 2). However, the favourable allele here originates from the Jandaroi parent rather than the Afghani accession, indicating that Jandaroi possesses other genes that are directly involved in the mechanism of Na⁺ exclusion.

It should be noted that Jandaroi was developed as the elite durum wheat cultivar adapted to the growth conditions prevalent in Australia; nevertheless, like other existing cultivated durum germplasms, Jandaroi, accumulates very high leaf Na⁺ concentrations, when compared to bread wheat [6, 8, 18]. This is typically because durum wheat do not contain an active native HKT gene involved in excluding Na⁺ from the shoot [6, 15, 16]. To date the only durum wheat with significant reductions in shoot Na⁺ have had HKT genes have had them introduced through selective processes, such as the incorporation of the Kna1 locus from the D genome of bread wheat [7, 15] or the Nax1 and Nax2 locus from T. monococcum [15, 17]. At this stage, there is no published information available that an HKT-like gene or any other potential candidate genes are located in the relevant genetic region of the chromosome 3BL. Therefore, a Jandaroi allele for Na⁺ exclusion with this region is unique. Consequently, further study is required to shed light on this particular locus, which seems to be very different from those in durum landraces.

Two loci affecting shoot Na⁺ concentration were identified in the F₂ segregating population, Jandaroi x AUS-14740, located on the long arms of chromosomes 4B and 3B. The loci need to be further fine mapped and candidate genes identified using SNP markers. This would facilitate the construction of a fine linkage map of the corresponding genetic regions. The HKT1;5-B1 gene identified in this study on the long arm of chromosome 4B requires further analysis as the most suitable candidate gene in both segregating populations originating from exotic Afghani landraces. Expression profiling of the Jandaroi and AUS-14740 alleles will indicate whether the difference in Na⁺ phenotype is due to differences in expression between the two accessions, while heterologous expression of the HKT1;5-B1 gene in yeast and Xenopus oocytes will enable characterisation of the transport properties of the protein. Marker assisted selection and backcrossing would facilitate the introduction of the beneficial allele of HKT1;5-B1 from AUS-14740 into Jandaroi, similar to the approaches used to introduce the Nax2 [17] and Kna1 [7] alleles into durum. Saline field trials would then determine the effect it has on yield. The relevant genetic region on the chromosome 3BL needs to be studied more extensively with a range of molecular markers in order to identify other candidate genes in the genetic region originating from cultivar Jandaroi.