Identification of Nicotiana tabacum Linkage Group Corresponding to the Q Chromosome Gene(s) Involved in Hybrid Lethality

Background A linkage map consisting of 24 linkage groups has been constructed using simple sequence repeat (SSR) markers in Nicotiana tabacum. However, chromosomal assignments of all linkage groups have not yet been made. The Q chromosome in N. tabacum encodes a gene or genes triggering hybrid lethality, a phenomenon that causes death of hybrids derived from some crosses. Methodology/Principal Findings We identified a linkage group corresponding to the Q chromosome using an interspecific cross between an N. tabacum monosomic line lacking the Q chromosome and N. africana. N. ingulba yielded inviable hybrids after crossing with N. tabacum. SSR markers on the identified linkage group were used to analyze hybrid lethality in this cross. The results implied that one or more genes on the Q chromosome are responsible for hybrid lethality in this cross. Furthermore, the gene(s) responsible for hybrid lethality in the cross N. tabacum × N. africana appear to be on the region of the Q chromosome to which SSR markers PT30342 and PT30365 map. Conclusions/Significance Linkage group 11 corresponded to the Q chromosome. We propose a new method to correlate linkage groups with chromosomes in N. tabacum.


Introduction
Hybrid lethality is a phenomenon that causes death of hybrids, and is one of the postzygotic mechanisms of reproductive isolation. This phenomenon, also called hybrid weakness or hybrid necrosis, is observed in plant species such as Nicotiana species [1], rice [2], wheat [3], Gossypium species [4] and Arabidopsis thaliana [5]. Hybrid lethality is considered to play a crucial role in the evolution of plants. An example supporting this hypothesis is that hybrid seedlings between two sympatric species, Papaver dubium and P. rhoeas, die, indicating that hybrid lethality functions as a barrier to gene flow between these species [6]. However, hybrid lethality can be an obstacle when desirable genes are to be introduced into cultivated species in breeding programs.
The genus Nicotiana includes 76 species classified into 13 sections [7]. Only N. tabacum (2n = 48, SSTT) and N. rustica are cultivated species, while the others are wild species. These cultivated species, especially N. tabacum, which belongs to section Nicotiana, have been crossed with many wild species. In several interspecific crosses, however, hybrid seedlings show hybrid lethality and die within a few weeks or months [1,8]. The hybrid lethality phenotype is specific to particular cross combinations in Nicotiana. Hybrid lethality in this genus is classified into five types based on the observed phenotype: Type I, browning of shoot apex and root tips; Type II, browning of hypocotyl and roots; Type III, yellowing of true leaves; Type IV, formation of multiple shoots; and Type V, fading of shoot color [1,9].
In previous studies, we focused on crosses between N. tabacum and species in Nicotiana section Suaveolentes. The section Suaveolentes includes 25 species restricted to Australasia and one African species, N. africana, which is the only known species in Africa [7]. These species are geographically isolated from the majority of species in other sections, most of which are distributed in the Americas. All species in section Suaveolentes are allotetraploids, and the section is considered to have originated from a single polyploidization event approximately 10 Mya, followed by speciation to produce the species known today [10]. Based on sequence analyses of nuclear-encoded, chloroplast-expressed glutamine synthetase (ncpGS), it has been suggested that the maternal progenitor is N. sylvestris (2n = 24, SS; section Sylvestres) and the paternal progenitor is Nicotiana section Trigonophyllae (2n = 24) [11]. N. tabacum is another allotetraploid that originated by interspecific hybridization of N. sylvestris with N. tomentosiformis (2n = 24, TT; section Tomentosae) with subsequent chromosome doubling approximately 200,000 years ago [10][11][12][13].
We crossed 11 species of section Suaveolentes with N. tabacum, and demonstrated that inviable hybrids were produced after crosses using nine species: N. africana, N. debneyi, N. excelsior, N. goodspeedii, N. gossei, N. maritima, N. megalosiphon, N. suaveolens and N. velutina [14][15][16]. The hybrid lethality observed in all these crosses at 28uC was identified as Type II, characterized by browning of hypocotyl and roots and suppression of symptoms at elevated temperatures ranging from 34 to 36uC. Our results also indicated that one or more genes on the Q chromosome in the S subgenome of N. tabacum are responsible for hybrid lethality in these crosses. Exceptionally, only N. benthamiana and N. fragrans, which belong to section Suaveolentes, produced 100% viable hybrids after crossing with N. tabacum [16,17].
Molecular markers would be useful to identify the location of the gene(s) responsible for hybrid lethality on the Q chromosome. Previously, we developed random amplified polymorphic DNA (RAPD) markers, inter-simple sequence repeat (ISSR) markers and sequence tagged site (STS) markers, all of which are specific to the Q chromosome [14,18]. However, the locations of these markers on the Q chromosome are unclear. Recently, the first linkage map of N. tabacum was constructed using simple sequence repeat (SSR) markers [19]. This map consists of 24 linkage groups, although three putative linkage groups (3a/3b, 8a/8b and 14a/14b) are each divided into two sublinkage groups. More recently, a high density linkage map containing 24 linkage groups has been developed using 2317 SSR markers [20]. Chromosomal assignments of the linkage groups have not yet been made, except for the A chromosome, as described in the Discussion.
Each chromosome of N. tabacum is lettered alphabetically (A to Z, excluding X and Y); chromosomes A to L belong to the T subgenome and M to Z to the S subgenome. A complete set of 24 monosomic lines of N. tabacum (Haplo-A to Z) has been established in the genetic background of 'Red Russian' [21,22]. When the above-mentioned nine species of section Suaveolentes are crossed with Haplo-Q, two types of hybrid seedlings are obtained: inviable hybrid seedlings possessing the Q chromosome and viable hybrid seedlings lacking the Q chromosome [14][15][16]. Using hybrids grown at 36uC, a linkage group corresponding to the Q chromosome can be identified from the linkage map of Bindler et al. [20].
In the present study, we identified a linkage group corresponding to the Q chromosome using hybrid seedlings from the cross Haplo-Q 6 N. africana by the Michelmore et al. method of bulked segregant analysis [23]. SSR markers on the identified linkage group were used to investigate Q chromosome involvement in hybrid lethality observed in the cross N. tabacum 6 N. ingulba. Furthermore, we estimated the location of the gene(s) triggering hybrid lethality using a Haplo-Q 6 N. africana hybrid with a possible deletion of the Q chromosome segment.

Preparation of DNA Bulks
Two types of DNA bulks, which differed in the presence or absence of the Q chromosome, were prepared using crosses N. tabacum 'Red Russian' 6 N. africana and N. tabacum Haplo-Q 6 N. africana. Five and 21 hybrid seedlings were obtained from F 1 seeds of the two respective crosses. All hybrid seedlings were cultured at 36uC, a temperature at which hybrid lethality is suppressed [16]. Four Q-chromosome-specific STS markers, QCS1, QCS2, QCS3 and QCS4, were used to determine whether the Q chromosome was present in these hybrid seedlings. All markers were detected in all five hybrid seedlings from the cross 'Red Russian' 6 N. africana and five of the hybrid seedlings from the cross Haplo-Q 6 N. africana, but not in the other 15 hybrid seedlings from that cross ( Table 1). The presence or absence of the Q chromosome was also confirmed by viability of the hybrid seedlings. When the hybrid seedlings cultured at 36uC were transferred to 28uC, the seedlings in which the markers were detected died, while those in which the markers were not detected survived. Two DNA bulks (Q+1 and Q+2) were prepared from hybrid seedlings possessing the Q chromosome and two DNA bulks (Q-1 and Q-2) were prepared from hybrid seedlings lacking the Q chromosome (Table 1).
Among 21 hybrid seedlings from the cross Haplo-Q 6 N. africana, we found one hybrid seedling in which QCS1, QCS3 and QCS4 were detected and QCS2 was not detected (Table 1). This hybrid seedling was viable after transfer to 28uC, suggesting that deletion of the Q chromosome region containing the gene(s) triggering hybrid lethality occurred. This hybrid seedling was used to determine the chromosomal region of the gene(s) triggering hybrid lethality as described below.

Identification of a Linkage Group Corresponding to the Q Chromosome
A linkage group corresponding to the Q chromosome was identified from a linkage map of N. tabacum [20]. First, we selected SSR markers at random from all linkage groups and checked whether they produced polymorphic bands between 'Red Russian' and N. africana. Most of the markers amplified one or two bands in each species, since both species are allotetraploid. Each band was named by the marker name followed by a letter. Many of these bands showed polymorphisms between 'Red Russian' and N. africana (Table 2). Next, we investigated whether the selected SSR markers produced polymorphic bands between the two types of DNA bulks. As a result, the SSR marker PT20383 on linkage group 11 was identified as showing polymorphism between the bulks (Table 2, Figure 1). Other SSR markers that mapped to linkage group 11 were investigated, and eight markers gave similar results to those of PT20383, indicating that linkage group 11 corresponds to the Q chromosome. This result was consistent with a report that the origin of markers in linkage group 11 is the S genome of N. sylvestris, an ancestor of N. tabacum [20,24].
Although the band termed PT1245A, detected by PT1245, showed polymorphism between 'Red Russian' and N. africana, the band showed no polymorphism between the two types of bulks (Table 2). Further studies using N. ingulba revealed that band PT1245A was not derived from the Q chromosome but another band, PT1245B, which showed no polymorphism between 'Red Russian' and N. africana, was derived from it (see below). Similarly, the polymorphic band PT1002A, detected by PT1002, showed no polymorphism between the two bulks ( Table 2). Although another band, PT1002B, showed no polymorphism between 'Red Russian' and N. africana, or between 'Red Russian' and N. ingulba, this band was likely derived from the Q chromosome.

Hybrid Lethality Observed in Crosses between N. tabacum and N. ingulba
Reciprocal crosses were carried out between 'Red Russian' and N. ingulba (section Suaveolentes) using conventional cross-pollination (Table 3). Although seeds were obtained from reciprocal crosses, the percentage of seed germination was significantly different: 82.5% when N. ingulba was used as the maternal parent and 5.7% when 'Red Russian' was used as the maternal parent. All hybrid seedlings from reciprocal crosses were inviable at 28uC (Figure 2). Characteristic symptoms of hybrid lethality were browning of the hypocotyl and roots, indicating that hybrid lethality in reciprocal crosses was Type II. The same results in reciprocal crosses suggested that hybrid lethality is due to the interaction of coexisting heterologous genomes, and not to a cytoplasmic effect.
To investigate whether hybrid lethality is suppressed at elevated temperatures, seven hybrid seedlings from the cross N. ingulba 6 'Red Russian' were obtained and cultured at 36uC. Hybrid lethality was completely suppressed at this temperature. However, when hybrid seedlings cultured at 36uC were transferred to 28uC, hybrid lethality was induced and all hybrid seedlings died (data not shown).

The Q Chromosome Causes Hybrid Lethality in Crosses between N. tabacum and N. ingulba
To determine whether the Q chromosome causes hybrid lethality in the cross N. tabacum 6 N. ingulba, N. ingulba was crossed with N. tabacum monosomic plants derived from the cross Haplo-Q 6 'Samsun NN' [18]. Those monosomic plants lacking the Q chromosome were used as maternal parents for the interspecific cross, since the transmission of the monosomic condition through pollen is very low [25].
Twenty-one flowers of monosomic plants were pollinated with N. ingulba. Four flowers produced capsules containing only powdery seeds and we were unable to obtain seeds with normal appearance. Since the percentage of flowers fertilized was low, we carried out test-tube pollination and ovule culture. Twenty placentas of monosomic plants were aseptically pollinated with N. ingulba. Sixteen hybrid seedlings were obtained from 189 ovules and cultured at 36uC. Four of these hybrid seedlings showed abnormal morphology and had no cotyledons and leaves. These four seedlings were excluded from further analysis.
Three Q-chromosome-specific STS markers, QCS2, QCS3 and QCS4, and 14 SSR markers mapped to linkage group 11 were analyzed. All STS and SSR markers excluding PT30024 produced polymorphic bands between parents, monosomic plants and N. ingulba (Table 4). When three of the STS and nine SSR markers that produced one polymorphic locus from monosomic plants were analyzed in hybrid seedlings, the bands detected by markers excluding PT1002 were observed in one hybrid seedling and not in the other 11 hybrid seedlings (Table 4, Figure 3). For three SSR markers, PT1245, PT30342 and PT30365, which produced two polymorphic bands from monosomic plants, one of the two bands was detected in one hybrid seedling and not in the other 11 hybrid seedlings, suggesting that the bands segregating in hybrid seedlings were specific to the Q chromosome (Table 4). Thus, using STS and SSR markers, 12 hybrid seedlings were divided into two types, one hybrid seedling possessing the Q chromosome and 11 hybrid seedlings lacking the Q chromosome.
Hybrid seedlings cultured at 36uC were transferred to 28uC. The one hybrid seedling possessing the Q chromosome died, while the 11 hybrid seedlings lacking the Q chromosome survived. This result indicated that one or more genes on the Q chromosome are responsible for hybrid lethality in the cross between N. tabacum and N. ingulba. Furthermore, we confirmed that linkage group 11 corresponds to the Q chromosome.

Dissection of a Haplo-Q 6 N. africana Hybrid with a Possible Deletion of the Q Chromosome Segment
To identify the possibly deleted region of the Q chromosome in a viable hybrid seedling from the cross Haplo-Q 6 N. africana (in which QCS1, QCS3 and QCS4 were detected and QCS2 was not detected), seven SSR markers showing polymorphism between the two types of bulks derived from hybrids of N. tabacum 6 N. africana (Table 2) were tested. Two markers, PT30342 and PT30365, were not detected, while the other five markers were detected (Table 5). This indicated that the chromosomal region where PT30342 and PT30365 are located was deleted in the hybrid seedling.

Discussion
Our previous studies revealed that many species of section Suaveolentes yield inviable hybrids after crosses with N. tabacum [14][15][16]. N. ingulba, which belongs to section Suaveolentes, is one species yielding inviable hybrids. In all these cases, hybrid lethality was Type II, which is suppressed at elevated temperatures ranging from 34 to 36uC, and the hybrid lethality gene(s) in N. tabacum were on the Q chromosome in the S subgenome. Recent studies based on internal transcribed spacer regions of rDNA, plastid genes, and ncpGS indicated that section Suaveolentes is a monophyletic group [11,12,26]. Based on this information, Suaveolentes species that yield inviable hybrids after crosses with N. tabacum should share the same gene triggering hybrid lethality by interaction with gene(s) on the Q chromosome. However, this conclusion is not applicable in all cases, since at least one species in section Suaveolentes, N. occidentalis, yields inviable hybrids even when they lack the Q chromosome [9].
The cross N. tabacum 6 N. africana rarely yields viable aneuploid hybrids or haploids [27,28]. In a previous study, we obtained a viable aneuploid hybrid lacking the Q chromosome from the reciprocal cross [16]. In addition to these aneuploids and haploids, we revealed that viable hybrids possessing this chromosome with a deletion of certain regions can be obtained from crosses. Although such deletions might be observed in all chromosomes of reciprocal hybrids between N. tabacum and N. africana, deletion within the Q chromosome would be preferentially detectable, since gene(s)  triggering hybrid lethality are located on the Q chromosome and most hybrids with an intact Q chromosome would die. A viable hybrid from the cross N. tabacum Haplo-Q 6 N. africana had a Q chromosome with a deletion of the region where PT30342 and PT30365 are located, suggesting that this region includes the causal gene(s) for hybrid lethality. This conclusion was drawn from only one hybrid plant with the deletion, since we could not obtain other deletion hybrids due to their rareness. Further studies will be needed to confirm and narrow down the position of the causal gene(s). A linkage map is important for mapping, marker-assisted selection and map-based cloning of desirable genes. In N. tabacum, the first linkage map using SSR markers has been constructed [19,20]. In the future, it will be desirable to assign linkage groups in the linkage map to each chromosome. For this purpose, monosomic lines should be useful, as demonstrated in the present study. Fortunately, a complete set of 24 monosomic lines of N. tabacum has been developed [21,22]. Therefore, hybrid lines lacking certain chromosomes can be obtained through crosses between monosomic lines and Nicotiana species. Nicotiana species other than N. tabacum might be desirable for the crosses, since many polymorphisms would be observed, as in the present study. However, in such cases, production of hybrid lines may be problematic due to crossing barriers in interspecific crosses. Interestingly, the two N. tabacum cultivars used for the construction of the linkage map were 'Red Russian' and 'Hicks Broad Leaf' [19,20]. Since the genetic background of all 24 monosomic lines of N. tabacum is 'Red Russian' [21,22], 'Hicks Broad Leaf' would be    the best parent to produce hybrid lines lacking a certain chromosome.
Another method to correlate linkage groups with chromosomes in N. tabacum has been reported by Vontimitta et al. [29], who mapped two genes responsible for the accumulation of cis-abienol and sucrose esters to a certain linkage group. Based on earlier monosomic analyses that revealed that both genes are located on the A chromosome [30,31], Vontimitta et al. determined the correlation between the linkage group and the A chromosome. This indirect method is useful to assign linkage groups to each chromosome. However, the number of genes allocated to the chromosome by monosomic analyses is too small to determine all chromosomal assignments of linkage groups. On the other hand, our method using hybrid lines lacking a certain chromosome is theoretically applicable to all chromosomes in N. tabacum. This is the first report of a practical method to correlate linkage groups with chromosomes in N. tabacum. 6 N. africana were obtained in a previous study [16] and used for bulked segregant analysis [23]. N. ingulba (2n = 40), which belongs to section Suaveolentes, was used for crosses with 'Red Russian' or F 1 progeny (2n = 47) derived from the cross Haplo-Q 6 N. tabacum 'Samsun NN'; the latter was identified by the Q-chromosomespecific STS marker QCS1 [18]. All plants used as parents were cultivated in a greenhouse.

Interspecific Crosses
Flowers of plants used as female parents were emasculated 1 day before anthesis and pollinated with pollen of plants used as male parents. F 1 seeds were sterilized with 5% sodium hypochlorite for 15 min. The sterilized seeds were sown in Petri dishes (60 mm diameter, 17 mm deep) containing 8 ml of 1/2 MS medium [32] supplemented with 1% sucrose and 0.2% Gelrite (pH 5.8) and then cultured at 28uC under continuous illumination (approximately 150 mmol m 22 s 21 ).
Test-tube pollination in combination with ovule culture was carried out as previously described [33]. Anthers of plants used as male parents were aseptically excised from still-closed flowers and stimulated to dehisce in an incubator (28uC). Flowers of plants used as female parents were emasculated 1 day before anthesis. On the next day, flowers of plants used as female parents were collected and their corolla, sepals and styles were removed. The ovaries were surface-sterilized with 70% ethanol for 30 s and with 5% sodium hypochlorite solution for 5 min. The ovary walls were peeled back to expose the placentas with intact ovules, and then the ovaries were placed in Petri dishes containing 8 ml of medium with 3% sucrose and 0.8% agar (pH 5.8). Pollen of plants used as male parents was spread on the surface of the placentas. Pollinated placentas were maintained at 28uC under continuous illumination. Fertilized and enlarged ovules were excised from placentas at 10 to 14 days after pollination and cultured in Petri dishes containing 8 ml of 1/2 MS medium supplemented with 3% sucrose and 0.8% agar (pH 5.8) at 28uC under continuous illumination.

Cultivation of Hybrid Seedlings
Hybrid seedlings were cultured at 28uC under continuous illumination. Some seedlings were transferred to flat-bottomed test tubes (25 mm diameter, 100 mm length) that contained 10 ml of 1/2 MS medium supplemented with 1% sucrose and 0.2% Gelrite (pH 5.8) immediately after germination and were cultured at 36uC under continuous illumination. Hybrid seedlings cultured at 36uC  Table 5. Analysis of a Haplo-Q 6 N. africana hybrid with a possible deletion of the Q chromosome segment using SSR markers on linkage group 11.
Forty-five SSR markers (Table 2) covering all linkage groups were detected as follows. Reaction mixtures contained 16 ThermoPol reaction buffer (New England Biolabs, Tokyo, Japan), 0.2 mM of each dNTP, 0.2 mM of each primer, 20 ng of template DNA, 0.5 U of Taq DNA polymerase (New England Biolabs) in a total volume of 20 ml. PCR amplification was performed using the TProfessional Basic Thermocycler (Biometra, Göttingen, Germany) programmed for 3 min at 94uC for initial denaturation, followed by 35 cycles of 30 s at 94uC, 30 s at 50 or 55uC, 1 min at 72uC, and a final extension of 5 min at 72uC. PCR products were separated by electrophoresis in a 3% agarose gel in TBE buffer and stained with ethidium bromide.