Identification of QTLs associated with tissue culture response of mature wheat embryos

Gene engineering technique is an important tool for functional studies of a given gene and genetic modification of cereal crops. An excellent tissue culture system is the prerequisite of highly efficient genetic transformation. Immature embryo is recognized as an excellent explant for wheat genetic transformation (Yin et al. 2011). However, obtaining immature embryos is limited by the growth season and controlled conditions, thus affecting the progress of tissue culture and therefore gene transformation (Medvecká and Harwood 2015). Thus, an alternative explant replacing immature embryos is favored. The mature embryo is such one of these alternatives as it is convenient to be obtained without limitation of growing seasons and development stages (Yin et al. 2011). Nonetheless, it is reported that the regeneration ability of the calli from mature embryos of wheat is limited, thus hindering its application. Numerous studies have focused on improving regeneration ability by adjusting constitutes and concentration of hormones (Gill et al. 2014; Li et al. 2013b; Murín et al. 2012; Shah et al. 2003; Turhan and Baser 2004), using whole explant culture, endosperm supported culture, thin pieces culture and explants cutting culture (Tao et al. 2008). It is reported that genotypes of the donors greatly affect the efficiency of callus induction since no culture conditions have been reported suitable for all genotypes (Bolibok and Rakoczy-Trojanowska 2006; Ge et al. 2006).

It is believed that genetic control of plant regeneration ability could be either qualitative (Reisch and Bingham 1980) or quantitative (Taguchi-Shiobara et al. 2006). A large number of studies have been reported on quantitative trait loci (QTLs) analysis for tissue culture response (TCR) in soybean (Yang et al. 2011), rice (Li et al. 2013a; Nishimura et al. 2005; Zhao et al. 2009), maize (Murigneux et al. 1994; Wan et al. 1992), barley (Bregitzer and Campbell 2001; Manninen 2000; Mano and Komatsuda 2002), rye (Bolibok et al. 2007), and wheat (Jia et al. 2007; Nielsen et al. 2015; Torp et al. 2001).

With introduced D genome of Aegilops tauschii, synthetic hexaploid wheat (SHW) could offer breeder novel genetic variations (Mares and Mrva 2008). It is widely accepted that SHW lines are superior in disease resistance, abiotic stress tolerance, grain quality, and tolerance to pre-harvest sprouting (Lage et al. 2003; Trethowan and Mujeeb-Kazi 2008; Yang et al. 2009). A full understanding of the genetic basis for these desirable traits is important in exploring favored genes/locus. The present study is one of such efforts aiming to identify important genes. Diversity Arrays Technology (DArT) is a rapid and DNA sequence-independent technique and can detect and type DNA variation at several hundred genomic loci in parallel (Jaccoud et al. 2001; Wenzl et al. 2004).

Compared with those in rice, fewer QTLs controlling TCR of mature embryo have been reported and our knowledge on the genetic basis of TCR in bread wheat is still limited. In this study, we are aiming at performing whole-genome linkage mapping using DArT markers to identify QTLs associated with callus induction from mature embryos and plantlet regeneration from calli using a population of recombinant inbred lines (RILs) derived from a cross between a synthetic hexaploid wheat (SHW-L1) and a cultivar Chuanmai 32.

Significant differences were not detected for callus rate between the two parents, SHW-L1 and Chuanmai 32 (Fig. 1). However, callus rate varied greatly among the RILs and showed a continuous distribution (p = 0.196 for test of normality) ranging from 12.2 to 96.2 % with a mean of 56.4 % (Table 1; Fig. 2). Significant difference in differentiation rate was also not detected between the two parents. The differentiation rate of the RIL population gave a continuous distribution (p = 0.413 for test of normality) with a range between 19.1 and 95.0 % with an average value of 68.1 %. For emergence rate, significant difference between parents was detected with SHW-L1 (17. 4 %) significantly higher than Chuanmai 32 (4.1 %) (Fig. 1). Emergence rate of the population (p = 0.061 for test of normality) ranges from 2.1 to 73.3 % with an average of 21.3 % (Table 1; Fig. 2). Pearson correlation analysis showed that callus rate was significantly and positively correlated with differentiation rate; and differentiation rate was significantly and positively correlated with emergence rate (Table 2).

Trait	Parentsa		2012b		2013b		2014b		Averagec
	SHW-L1	Chuanmai 32	Mean	Range	Mean	Range	Mean	Range	Mean	Range
Callus rate	59.3	62.2	57.8	14.2–92.5	48.9	4.5–99.0	62.4	18.0–97.0	56.4	12.2–96.2
Differentiation rate	43.2	47.3	75.4	16.3–100.0	39.3	3.3–85.0	89.4	37.5–100.0	68.1	19.1–95.0
Emergence rate	17.4**	4.1	13.3	0.0–70.0	30	0.0–80.0	–	–	21.3	2.1–73.3

** Significant differences between SHW-L1 and Chuanmai 32 at p = 0.01

^aThe values were means of data from 3 or 2 years

^bThe values were calculated from the data of the 171 lines of the population

^cThe average was calculated by the data of 2 or 3 years

	Callus rate	Differentiation rate	Emergence rate
Callus rate
Pearson correlation	1	0.315**	0.066
Significance (two-tailed)		<0.01	0.43
# of analyzed lines	149	149	149
Differentiation rate
Pearson correlation	0.315**	1	0.318**
Significance (two-tailed)	<0.01		<0.01
# of analyzed lines	149	149	149
Emergence rate
Pearson correlation	0.066	0.318**	1
Significance (two-tailed)	0.43	<0.01
# of analyzed lines	149	149	149

**Significant differences at p = 0.01

Three QTLs for callus rate were identified in each of the three seasons assessed. They were located on chromosomes 1D, 5A, and 6D, respectively (Table 3; Fig. 3). Each of these loci explained between 10.16 and 11.82 % of the phenotypic variation. Positive alleles of these QTLs were derived from Chuanmai 32 (Table 3). A single QTL, QDiffr.sau-1A, was detected for differentiation rate only in a single growing season. The phenotypic variation explained by this QTL was 10.96 %. Positive allele of this locus was derived from SHW-L1. Two QTLs for emergence rate were identified in a single growing season and they were located on 3B and 4A, respectively. The positive alleles of these QTLs were derived from Chuanmai 32. The phenotypic variation for these QTLs was 9.88 and 10.30 %, respectively (Table 3).

Trait	QTL	Chromo-some	Trail	Position	Left markera	Right markera	LODb	PVE (%)c	Addd
Callus rate	QCallr.sau-6D	6D	2012	1.21	wPt-730539	rPt-7068	3.88	11.44	−0.07
	QCallr.sau-5A	5A	2013	54.41	wPt-2873/wPt-1200	wPt-733461	3.84	10.16	−0.09
	QCallr.sau-1D	1D	2014	55.11	wPt-730941	wPt-741323	5.02	11.82	−0.07
Differentiation rate	QDiffr.sau-1A	1A	2012	180.71	wPt-3698/wPt-730408	wPt-666607/wPt-665725	3.96	10.96	0.07
Emergence rate	QEmr.sau-3B	3B	2014	78.71	wPt-1191	wPt-744159	3.50	9.88	−0.05
	QEmr.sau-4A	4A	2014	21.61	wPt-0798	wPt-0764/wPt-8657/wPt-734161	3.50	10.3	−0.06

^aThe left and right flanking marker of the interval of LOD peak value for a given QTL

^bThe logarithm of odds

^cPercentage of the phenotype variation explained

^dAdditive effect at QTL, positive values indicate the effect from SHW-L1, negative values indicate the effect from Chuanmai 32

As indicated in the study of Jia et al. (2007), group 5 chromosomes are likely to play key roles in TCR of Triticeae crops. Here, a QTL QCallr.sau-5A for callus rate was identified. Previous studies identified QTLs for percentage of embryos forming a callus (PEFC) and percentage of calli regenerating plantlets (PCRP) on 5A and 5B chromosomes during culturing of both mature and immature wheat embryos (Jia et al. 2007, 2009). In addition, association analysis identified loci on 5A (wPt-6135) and 5B (tPt-4184) associated with embryo formation in culturing of wheat microspores (Nielsen et al. 2015). In the culturing of barley immature embryo, a QTL controlling callus growth was also identified on 5H in barley (Mano and Komatsuda 2002).

In correspondence to QCallr.sau-1D for callus rate and QDiffr.sau-1A for differentiation rate detected on 1D and 1A in wheat, a QTL controlling shoot differentiation ratio was identified on 1H chromosome in barley (Mano and Komatsuda 2002). Nielsen et al. (2015) identified a locus on 1B which was associated with albino and green plantlet regeneration. Different from QTLs for PCRP in wheat mature and immature embryo culture located on 3A (Jia et al. 2007, 2009), a QTL QEmr.sau-3B for emergence rate was identified on 3B. A QTL controlling shoot differentiation ratio was detected on 3H chromosome for immature embryo culture in barley (Mano and Komatsuda 2002). These results indicated that both group 1 and group 3 chromosomes of Triticeae crops could be important in TCR.

In addition to those QTLs on group 1, 3, and 5 chromosomes, two QTLs for callus and emergence rates were identified on located on 6D and 4A, respectively. The fact that very few QTLs for TCR in wheat were identified on 6D and 4A indicates that these two QTLs could be novel loci.

Some QTLs described in this study, were detected in one environment only, indicating that TCR is dependent on not only genotype but also environment. This result is in accordance with the study showing that green plantlet formation is highly dependent on both environment and genotype during wheat microspore culture (Holme et al. 1999). Thus, it is recommended that the wheat should be planted in controlled environments such as glasshouse for harvesting mature seeds for tissue culture.

Here, we report the identification of QTLs for callus induction and callus regeneration from mature embryos based on a RIL population. Three QTLs for callus rate were identified and they were located on chromosomes 1D, 5A, and 6D, respectively, and they explained phenotypic variation ranging from 10.16 to 11.82 %. One QTL for differentiation rate was detected only with 10.96 % of the phenotypic variation explained. Two QTLs for emergence rate were identified and they were located on 3B and 4A, respectively, with 9.88 and 10.30 % of phenotypic variation.