¹University of Adelaide, Dept. of Plant Science, PMB 1, Glen Osmond, SA, 5064, Australia

Introduction

Breeders are constantly looking for new and innovative methods of fast tracking the selection process used for the development of new barley varieties. One such method is marker assisted selection (MAS). MAS allows complex traits to be screened in the early generations of a breeding program. This is usually much before the standard methods of phenotyping can be employed, particularly in the case of malting quality traits.

Barr et al. (2000) identified nine key steps in the identification of molecular markers linked to traits for use in MAS. The last of these steps was the validation of putative markers in alternative genetic backgrounds. This step is particularly important when markers have been found using QTL mapping as their location is often poorly defined.

The Australian national barley molecular marker program (NBMMP) and an Adelaide University research team have produced linkage maps of a number of populations which have been used to identify quantitative trait loci (QTL) associated with several complex malting quality traits in barley (Table 1). However, before markers associated with these QTL can be used for MAS they must be validated in alternative genetic backgrounds.

This paper reports studies on the validation of markers associated with key quality traits in populations genetically distant from the original mapping populations. Details of relevant QTL and associated markers are provided in Table 1. The chromosomal location of the markers used in the study is shown schematically in Figure 1.

Table 1: The markers used in the study to validate regions found to be associated with HWE, DP, alpha-amylase and beta-amylase in Australian mapping populations.

Trait#	Population from which QTL were identified	Parent donating higher allele	Chrom	Marker used in validation
HWE	Chebec x Harrington1	Chebec	1H	Xbcd508
HWE	Chebec x Harrington1	Harrington	5H	Xabg57, XGMS01
HWE	Galleon x Haruna nijo2	Haruna nijo	2H	Xpsr108
HWE	Alexis x Sloop, Alexis x Sloop-sib1	Alexis	2H	Xabg14
HWE	Alexis x Sloop, Alexis x Sloop-sib1	Alexis	5H	Xabg712
HWE	Alexis x Sloop, Alexis x Sloop-sib1	Alexis	5H	Xabc310
HWE, DP, alpha, beta	Alexis x Sloop, Alexis x Sloop-sib1	Alexis	1H	XEbmac501
DP, alpha, beta	Alexis x Sloop, Alexis x Sloop-sib1	Sloop, Sloop-sib	4H	Xcdo669
DP, alpha, beta	Alexis x Sloop, Alexis x Sloop-sib1	Sloop, Sloop-sib	5H	Xbmac113
DP, alpha, beta	Amaji nijo x WI25853	Amaji nijo	5H	*Xxabg705, Xabg3

*Flanking markers, ^#HWE: hot water extract, DP: diastatic power, alpha: α-amylase, beta: β-amylase, ¹NBMMP, unpublished, ² Collins et al., 1999, ³Asayama, unpublished

Materials and Methods

Details on the six populations chosen for validation purposes, including parents, number of lines, and number of years and sites grown, are provided in Table 2. Harvested samples of each line from each population were malted in a Phoenix Automated Micromalting system at 15°C using the malting schedule described in the Waite Barley Quality Evaluation Laboratory Barley Quality Report (WBQEL, 1997). Hot water extract (HWE) was measured using small scale versions of the EBC method (EBC Analytica, 1998). Diastatic power (DP), alpha amylase (alpha) and beta amylase (beta) were measured as described in the WBQEL Barley Quality Report (1997).

The markers shown in Table 1 and schematically represented in Figure 1 were either selected on the basis of their proximity to the peak of the QTL or as flanking markers to the QTL region. Microsatellite markers XGms01 and RFLP marker xabg57 were both associated with a QTL on chromosome 5H with equal statistical significance but were both tested to investigate the potential of different marker systems for this region. DNA extraction and RFLP analysis followed general methods. Microsatellite analysis followed the method of Karakousis et al. (2000). Initially the parents of each population were screened for marker allele polymorphisms.

The lines of each population were separated into two groups, consisting of those carrying the high malt quality allele and those carrying the low quality allele, omitting heterozygotes. Least square means were calculated from an ANOVA using the marker allele groups as the factor. Statistical analysis was performed using JMP (v3.1.6, SAS Institute Inc, 1995).

Figure 1: A schematic diagram of the chromosomal location of markers found to be associated with a number of malting quality traits in Australian mapping populations. ^aKarakousis et al., 2001, ^bLangridge et al., 1995, ^cNBMMP, unpublished

Table 2: The mean HWE, mean DP, mean alpha amylase and mean beta amylase results for the populations used to validate QTL.

Population	No.	Year	Site	Mean HWE	Mean DP	Mean alpha	Mean beta
Barque/Harrington	61	1996	Charlick	78.9
Barque/Haruna nijo	45	1996	Charlick	78.3
Sloop/Harrington//VB9623	51	1998	Charlick	77.6	780	110	670
Sloop/Harrington//VB9624	44	1998	Charlick	79.3	571	106	465
Sloop-sib/Alexis//VB9624	45	1998	Charlick	77.6	569	94	475
DH115/Sloop-sib//Amaji nijo/Alexis	89	1998	Charlick	77.2	530	74	455

Results and Discussion

Hot Water Extract

Table 3 shows the mean HWE results for the two allele groups at four marker loci in four populations containing the parents Harrington and Haruna nijo. A highly significant three percent difference (P<0.001) in HWE was observed between lines carrying the Haruna nijo and Barque alleles at the Xpsr108 locus (Table 3). Lines carrying the Harrington allele at the Xpsr108 locus produced significantly (P<0.05) higher HWE than lines carrying the Barque allele despite the fact that no significant HWE QTL in this region was identified in the Chebec x Harrington mapping population. The malt quality and alternative allele groups were found not to be significantly different for HWE in the other two populations, which also contained Harrington as a parent.

The markers Xabg57 and Xgms01 are located within the region of chromosome 5H found to be significantly associated with HWE in the mapping populations of Chebec x Harrington (ref) and Harrington x TR306 (Mather et al., 1997). To validate markers in this region, three populations containing Harrington were probed with the marker Xabg57. Two of these populations were also screened with the marker Xgms01 to assess how robust this QTL is when using different marker systems. A significant difference (P<0.05, P<0.01) in HWE between lines carrying the Harrington and alternative allele for Xabg57 was identified in the Barque/Harrington and Sloop/Harrington//VB9623 populations. A similar result eventuated from using Xgms01 in the Sloop/Harrington//VB9623 population. However in the third population, Sloop/Harrington//VB9624, lines carrying the Harrington allele were not significant different for HWE from lines carrying the alternative allele. Lines in the population Barque x Haruna nijo were also screened with the marker Xabg57 but there was no significantly difference in HWE between allele classes.

The influence of this locus on the DP enzymes was previously investigated in a number of breeding populations including two of the populations used in this investigation, Barque/Harrington and Barque/Haruna nijo (Coventry et al., 1999). Lines carrying the malting quality parent allele at this locus produced significantly higher alpha-amylase than lines carrying alternative alleles. This indicates that this region is not only influencing malt extract but also other important malt quality traits.

The final marker investigated with these populations was Xbcd508 on chromosome 1H. This region was found to be associated with HWE in the Chebec x Harrington (NBMMP, unpublished) and Harrington x TR306 (Mather et al., 1997) populations. In both cases Harrington allele was associated with lower HWE. Despite this there was no significant difference between lines carrying the Harrington allele and lines carrying the alternative allele in the three populations screened. In the case of the Sloop/Harrington//VB9623 population, there were no lines found to be carrying the Harrington allele. It is possible that a QTL controlling an important agronomic trait is co-located in this region and the lines carrying that particular plant type were culled out before the population was selected for this experiment.

The interaction effects on HWE between the two marker loci Xabg57and Xpsr108 were also investigated (Table 4). The lines from the four populations were divided into the four alternative allele groups based on the two markers. In the population Barque/Harrington, the lines carrying the Harrington allele at both marker loci produced significantly (P<0.01) higher HWE then all other groups. Lines carrying Harrington alleles at both loci produced 0.9% higher HWE than those carrying the Harrington allele at Xabg57only in the Barque/Harrington population and 0.2% in the Sloop/Harrington//VB9623. These results were not confirmed in the other two populations containing Harrington as a parent. The lines carrying the Harrington alleles at both marker loci in the later population were not significantly higher then the lines without Harrington alleles at both loci.

The third population Sloop-sib/Harrington//VB9624 only had one line that had the Harrington allele had both marker loci (Table 4). Hence, this line could not be statistically separated from the other groups, even though it had a high extract of 82%.

Lines carrying Haruna nijo alleles at both Xabg57and Xpsr108 loci in the population Barque x Haruna nijo were significantly higher (P<0.05) for HWE than lines carrying Barque alleles at both loci (Table 4). Lines carrying Harrington alleles at Xabg57 only, however, were not significantly different from those carrying Barque alleles at both loci. Therefore there would be no reason to use the marker Xabg57 in this population, as it would give no added benefit to selection.

Table 3: The mean HWE results (% dry basis) of the different allele classes at four marker loci in four populations containing Harrington or Haruna nijo as a parent

		Xpsr108 (2H)		Xabg57 (5H)		Xgms01 (5H)		Xbcd508 (1H)
Population	allele	aHigh HWE	Other	aHigh HWE	Other	aHigh HWE	Other	aHigh HWE	Other
Barque/Harrington	HWE	79.2*	78.4	79.5*	78.7			79.1	78.7
	No.	29	26	20	41			27	30
	SD	1.5	1.4	1.5	1.3			1.7	1.3
Barque x Haruna nijo	HWE	80.4***	77.1	77.9	78.8
	No.	14	28	28	17
	SD	1.7	2.1	2.5	2.3
Sloop/Harrington//VB9623	HWE	77.5	77.6	78.6**	77.1	78.8**	77.1	-	77.6
	No.	30	21	14	34	13	28	0	51
	SD	1.7	1.4	1.4	1.5	1.2	1.4		1.6
Sloop/Harrington//VB9624	HWE	79.4	79.2	79.6	79.1	79.6	79.1	79.4	79.1
	No.	6	38	15	25	16	24	17	25
	SD	1.6	1.2	1.4	1.1	1.4	1.2	1.1	1.3

* significantly different (P<0.05), ** significantly different (P<0.01), * *significantly different (P<0.001)

^aHarrington or Haruna nijo allele, SD: Standard deviation, No: number lines carrying the allele

Four markers were validated in populations that contained Alexis as a parent (Table 5). The first of these was Xabg14 located on chromosome 2H. Lines carrying an Alexis allele at Xabg14 produced 1% higher HWE (P<0.05) than lines carrying alternatives alleles in the population Sloop-sib/Alexis//VB9624. There was no significant difference, however, between allele classes at Xabg14 in the DH115/Sloop-sib//Amaji nijo/Alexis population.

Two markers were validated for HWE on chromosome 5H, namely Xabg712 and Xabc310. There was only a polymorphism among lines in one of the populations, DH115/Sloop-sib//Amaji nijo/Alexis at the Xabg712 locus (Table 5). There was no significant difference, however, between the two allele classes at this marker locus. Lines carrying an Alexis allele at Xabc310 in the DH115/Sloop-sib//Amaji nijo/Alexis population produced significantly higher HWE than those carrying alternative alleles. The interaction effects between both marker loci were also investigated (Table 6). While the HWE mean of the lines carrying the Alexis allele at both marker loci was significantly higher (P<0.05) than the mean of the lines not carrying the Alexis allele, the HWE was not significantly different to the mean of the lines carrying an Alexis allele at the Xabc310 only. This indicates there is no benefit in using both of these markers to select for HWE.

The final marker validated was XEbmac501 located on chromosome 1H (Table 5). There was no polymorphism in the population Sloop-sib/Alexis//VB9624 for this marker. The mean HWE for the two allele classes, Alexis and alternative alleles, at this marker, in the population DH115/Sloop-sib//Amaji nijo/Alexis, were not significantly different.

The interaction effects between the two marker loci, Xabg14 and Xabc310 were also assessed in the two Alexis derived populations (Table 7). The population Sloop-sib/Alexis//VB9624 contained no lines that carried the Alexis allele for both markers. In the population DH115/Sloop-sib//Amaji nijo/Alexis the mean HWE for the lines with the Alexis allele at both marker loci was observed to be 1.4% higher then the mean of the result for the lines without the Alexis allele at both marker loci. However due to the small number of lines carrying the Alexis allele at both markers this was not a statistically significant different.

Table 4: The mean HWE results (% dry basis) for lines carrying the high (High) malting quality or alternative (other) parent allele for Xpsr108 and Xabg57 in four populations containing Harrington or Haruna nijo as a parents.

allele type		Barque/Harrington			Barque/Haruna nijo			Sloop-sib/Harrington// VB9623			Sloop-sib/Harrington// VB9624
Xpsr108 (2H)	Xabg57 (5H)	No.	HWE	SD	No.	HWE	SD	No.	HWE	SD	No.	HWE	SD
other	other	15	78.3	1.5	12	78.0bd	1.9	14	77.1	1.3	13	79.3	0.9
other	High	10	78.7	1.2	16	76.5ef	2.0	7	78.3	1.3	15	79.4	1.3
High	other	19	78.7	1.2	5	80.7be	2.0	20	77.0d	1.6	4	78.8	1.2
High	High	8	80.6a	1.3	9	80.1df	1.6	7	78.8d	1.5	1	82.0	-

^asignificantly different (P<0.01) to all groups for that population, ^bdalleles with letters in common are significantly different (P<0.05) for that population, ^efalleles with letters in common are significantly different (P<0.001) for that population, SD: Standard deviation, No: number lines carrying the allele

Table 5: The mean HWE results (% dry basis) of the different allele classes at four marker loci in two populations containing Alexis as a parent.

		Xabg14 (2H)		Xabg712 (5H)		Xabc310 (5H)		XEbmac501 (1H)
Population	allele	Alexis	Other	Alexis	Other	Alexis	Other	Alexis	Other
Sloop-sib/Alexis// VB9624	HWE	78.4*	77.4	np		78.3	77.4	np
	No.	11	33			9	36
	SD	1.1	1.4			1.5	1.3
DH115/Sloop-sib// Amaji nijo/Alexis	HWE	77.6	76.9	77.3	77	77.9*	76.9	77.5	77
	No.	18	64	37	48	14	57	14	71
	SD	1.1	1.4	1.3	1.4	1.1	1.5	1.7	1.3

*significantly different (P<0.05), SD: Standard deviation, No: number lines carrying the allele, np: no polymorphism

DP, alpha-amylase and beta-amylase

Three regions in the mapping population Alexis x Sloop were found to be associated with DP, alpha-amylase and beta-amylase. The markers most significantly associated with these traits were Xbmac113, Xcdo669 and XEbmac501. Polymorphisms for the marker Xcdo669 could not be found in the two populations investigated. Likewise a polymorphism for the marker XEbmac501 was not found in the population Sloop-sib/Alexis //VB9624. However at this marker locus the Alexis allele was associated with significantly higher (P<0.05 and P<0.01) values for all three traits in the other population, (DH115 x Sloop-sib) x (Amaji nijo x Alexis). The third marker screened was Xbmac113 on chromosome 5H. No significant difference could be found between the two allele classes at this marker locus in either of the populations.

A single region, associated with DP was found in the mapping population, Amaji nijo x WI2585, on chromosome 5H (Asayama, unpublished). This region is flanked by the markers Xabg3 and Xabg705. When the effect of different alleles was examined at these two marker loci in the population DH115/Sloop-sib//Amaji nijo/Alexis, neither was significantly different (Table 8) for ant of the three traits. Table 9 shows the interaction effects between the two flanking markers. While having the Amaji nijo allele at both loci increased the alpha activity by 17%, this increase was not significant.

Table 6: The effect of the alternate alleles on the mean HWE results (% dry basis) at the two markers Xabc310 and Xabg712 in the DH115/Sloop-sib//Amaji nijo/Alexis population.

Xabc310	Xabg712	No.	HWE	SD
Other	Other	35	76.7a	1.5
Other	Alexis	22	77.2	1.4
Alexis	Other	4	78.0	0.4
Alexis	Alexis	10	77.9a	1.3

^asignificantly different (P<0.05)

Table 7: The effect of the alternate alleles on the mean HWE results (% dry basis) at the two markers Xabc310 and Xabg14 in two populations containing Alexis.

allele type		DH115/Sloop-sib//Amaji nijo/Alexis			Sloop-sib/Alexis//VB9624
Xabg14	Xabc310	No.	HWE	SD	No.	HWE	SD
Other	Other	47	76.7	1.5	24	77.1a	1.3
Other	Alexis	10	77.8	1.3	9	78.3	1.5
Alexis	Other	9	77.2	1.2	11	78.3	1.1
Alexis	Alexis	4	78.1	0.5	0	-	-

^asignificantly different (P<0.05) to the other groups

Table 8: The mean DP, alpha-amylase and beta-amylase results (μmoles maltose equiv) of the different allelic groups at four markers in two populations containing Alexis as a parent.

		Xbmac113 (5H)		Xcdo669 (4H)		XEbmac501 (1H)		Xabg3 (5H)		Xabg705 (5H)
Population	allele	Sloop	Other	Sloop	Other	Alexis	Other	Amaji nijo	Other	Amaji nijo	Other
Sloop-sib/ Alexis//VB9624	DP	579	564	np		np
	alpha	99	92
	beta	480	472
	No.	15	30
DH115/Sloop-sib// Amaji nijo/Alexis	DP	524	533	np		601*	515	544	522	525	531
	alpha	73	75			86**	72	80	71	76	73
	beta	451	458			516*	443	467	450	449	456
	No.	14	69			14	71	26	57	30	53

*significantly different (P<0.05), **significantly different (P<0.01), No: number lines carrying the allele, np: no polymorphism

Table 9: The mean DP, alpha-amylase and beta-amylase results (μmoles maltose equiv) obtained by dividing the DH115/Sloop-sib//Amaji nijo/Alexis population into the allelic groups at the markers Xabg705 and Xabg3.

allele type		No.	DP	SD	alpha	SD	beta	SD
Xabg705	Xabg3
other	other	39	521	122	71	19	450	119
other	Amaji nijo	14	555	111	78	18	477	101
Amaji nijo	other	18	521	136	72	18	449	129
Amaji nijo	Amaji nijo	11	537	123	83	13	454	115

Conclusion

A summary of the marker loci examined for each trait is shown in Table 10. In an attempt to assess the impact of each locus, Table 10 provides an indication of the response to selection for the allele donated from the high malt quality parent. Overall there were ten cases where the allele from the high quality parent produced a significant positive effect while there was no significant negative effect associated with any marker locus assessed. It would appear that the risks, in terms of malt quality, of selecting for any of the regions assessed would be low and there would most likely to be a positive overall effect on malt quality. A practical strategy for the implementation of MAS for improved malt quality, therefore, would be to genotype progeny at all of these key loci and select for the allele from the high quality parent at all loci.

Table 10: The response to selection from choosing the positive allele from five high malting quality donors when tested in six different validation populations.

HWE

DP

alpha-amylase

beta-amylase

Chrom

marker locus

Harr

HN

Alexis

Alexis

AN

Sloop

Alexis

AN

Sloop

Alexis

AN

Sloop

1H

XEbmac501

(+)

+

++

+

Xbcd508

(+),(+)

2H

Xpsr108

+, =, =

+++

Xabg14

+, (+)

5H

Xbmac113

(+), (-)

(+), (-)

(+), (-)

Xabg705

(-)

(+)

(-)

Xabg3

(+)

(+)

(+)

Xabg712

(+)

Xabc310

(+), +

Xabg57

+,++, (+)

(-)

Xgms01

++,(+)

+++ significant P<0.001, ++ significant P<0.01 + significant P<0.05, (+) not significant but positive,

= not significant, (-) not significant but negative

Acknowledgements

Barrett Burston Maltings, Joe White Malting Co. and The University of Adelaide fund this work. The authors wish to thank the members of the Waite Barley Improvement Program, the CRC for Molecular Plant Breeding and National Barley Molecular Marker for their assistance. Kerrie Willsmore and Elysia Vassos are thanked for their help with enzyme analysis.

References

1. Barr, A.R., Jefferies, S.P., Warner, P., Moody, D.B., Chalmers, K., and Langridge, P. (2000) Barley Genetics VIII, 8^th Barley Genetics Symposium, Volume I, 167-178

2. Collins, H.M., Logue, S.J., Jefferies, S.P., Stuart, I.M., and Barr, A.R. (1999) Proceedings of the 9th Australian Barley Technical Symposium, Melbourne, Australia, 2.44.1-2.44.6

3. Coventry, S., Barr, A.R., Jefferies, S.P., Chalmers, K., Logue, S.J. and Langridge, P. (1999) Proceedings of the 9th Australian Barley Technical Symposium, Melbourne, Australia, 3.16.1-3.16.5

4. European Brewery Convention Analytica (EBC), (1998) Verlag Hans Carl Geranke-Fachverlag 4.4

5. Karakousis, A., Chalmers, K., Barr, A.R., and Langridge, P. (2000) Barley Genetics VIII, 8^th Barley Genetics Symposium, Volume III, 64:66

6. Karakousis, A., Gustafson, J.P. and Chalmers, K.J., (2001) Plant and Animal Genome IX Conference, Poster SSR

7. Langridge, P., Karakousis, A., Collins, N., Kretschmer, J. and Manning, S. (1995) Molecular. Breed., 1:389-395

8. WBCEL, The Waite Barley Quality Evaluation Laboratory Barley Quality Report (1997)

9. Mather DE, Tinker NA, LaBerge DE, Edney M, Jones BG, Rossnagel BG, Legge WG, Briggs KG, Irvine RB, Falk DE and Kasha KJ, (1997), Crop Science, 37, 544-554