INTRODUCTION

The stem rust race Ug99 together with its newest variants and partially characterized Ethiopian races threatens durum wheat production in Ethiopia and, more in general, in the West Asia and North Africa (WANA)

region, where durum wheat is a primary cereal crop (Pretorius et al., 2000; Jin et al., 2008, 2009). Association mapping (AM) on germplasm collections has been recently used to validate the effects of genes/QTLs

previously discovered with bi-parental mapping and/or to discover new useful allelic variants (Buckler and Thornsberry, 2002; Maccaferri et al., 2010). Linkage disequilibrium (LD) in the cultivated durum wheat

germplasm ranges from 5 to 10 cM (Maccaferri et al., 2005, Somers et al., 2007), thus enabling a whole-genome scan analysis for marker-trait associations with a relatively low number of markers as compared to

other species characterized by higher LD decay rates. So far, the main locus for stem rust (Sr) resistance (Sr13) in durum wheat has been mapped on chr. 6A (Simons et al., 2010). In this project, a panel of 183 elite

durum wheat cultivars has been tested for stem rust (Puccinia graminis) response at adult stage to identify loci affecting resistance to Ug99 and other stem rust races.

Searching via association mapping for novel sources of resistance to Ug99 and other Ethiopian stem rust races in durum wheat

Tesfaye L. Dugo1,4, Marco Maccaferri1, Ayele Badebo2, Karim Ammar3, Maria C. Sanguineti1, Jose Crossa3 and Roberto Tuberosa1

MATERIALS AND METHODS

Association panel and stem rust response evaluation

A germplasm collection (Durum Panel) suitable for AM studies and representing most of the major durum

breeding pools has been assembled at DiSTA (University of Bologna, Italy; Maccaferri et al., 2006). One

hundred eighty-three lines of the durum panel along with three hallmark check cultivars (Altar 84, Yavaros

79 and Simeto) were evaluated at Debre Zeit Research Station (Ethiopia) in observation plots (un replicated

design) for four consecutive growing seasons (off-season and main-season) in 2009 and 2010 under

artificial inoculation with Ug99 (TTKSK race) and a mixture of durum-specific races prevalent in Ethiopia.

Maximum Disease Severity Scores (DSS) as percentage of stem covered by pustules were visually

recorded at the epidemic peaks and converted into Coefficient of Infection (Stubbs et al.,1986) for

subsequent analysis. ANOVA was carried out considering seasons as replicates and using the genotype ×

season interaction as error variance.

Molecular markers and association mapping analysis

The durum panel has been profiled with 320 SSRs markers and 876 DArT markers of known genetic map

position as determined from a joint analysis of segregation data from four recombinant inbred mapping

populations (Maccaferri et al., unpublished results). Two STS markers were also included for profiling the

panel. The TASSEL 2.1 program based on a mixed linear model with population structure controlled by the

kinship matrix (K) and Q matrix (Q) was used for detecting marker-trait associations. Only allelic variants

with frequency higher than 0.10 were used in the association test with the phenotypic data. Significant

markers were assigned to a QTL if the distance between the significant marker and the most significant

marker in the QTL region was less than 5 cM, or if the significant marker was less than 20 cM away but in

very high LD (R2 > 0.6) with markers in the QTL region (Massman et al., 2011). The length of the QTL

region was then based on the map position of all included markers.

References

Buckler and Thornsberry 2002. COPB 5: 107-111

Jin et al. 2008. Plant Dis 92: 923-926

Jin et al. 2009. Plant Dis 93: 367-370

Maccaferri et al. 2005. Mol Breed 15: 271-289

Maccaferri et al. 2006.Plant Genet. Res. 4: 79-85

Maccaferri et al. 2010. Mol Breed 26: 189-228

1DiSTA, University of Bologna, Viale Fanin 44, 40127 Bologna, Italy, 2Debre Zeit Agricultural Research Center, Debre Zeit, Ethiopia, 3CIMMYT, Int.

Apdo Postal 6-641, Mexico D.F., Mexico 06600, 4Sinana Agricultural Research Center, Bale-Robe, Ethiopia

Massman et al. 2011. Mol Breed 27: 439-454

Pretorius et al. 2000. Plant Dis 84: 203

Simons et al. 2010. TAG: DOI 10.1007s00122-010-14

Somers et al. 2007. Genome 50: 557-567

Stubbs et al. 1986. Cereal Dis. Method. Manual, pp:

45-46

RESULTS

Phenotypic evaluation for stem rust response

On average, the Coefficient of Infection (CI) in the off-seasons of both years was markedly higher than that

registered in the main seasons with mean values of 41.7-48.7 and 28.9-33.2, respectively (Table 1). The CI

frequency distributions indicates (data not shown) a continuous variation in response to the disease

suggesting that this trait in the durum panel is controlled by several genes/QTLs. The heritability coefficient

(calculated across environments using data from un replicated experiments, hence likely overestimated)

was equal to 0.80. About 13 lines had CI values < 10 in both off-seasons while more than 30 lines had CI

value < 10 in both main seasons.

Association mapping results

The QTLs with consistent effects across three to four seasons are reporter in Table 2. These QTLs were

those with the highest overall R2 values based on the season-mean data (in most cases comprised between

4.0 and 7.0%) as well as for single seasons (values ranging from 1.1 to 11.3%).In particular, regions on

chromosomes 1BS (QTL-tagging marker barc8), 2AS (gwm1045), 3AS (wPt-7972), 6AL (gwm427 and

CD926040) and 7AS (wPt-2799) showed the highest R2 values and all these QTLs were tagged by a series

of adjacent markers that supported the primary QTL effect. Regions on chromosomes 2BL, 3AL and 5AL had

consistently high R2 values but were identified by single markers.

A comparative analysis of what has so far been published on Sr loci in wheat indicates the overlap of the chr.

6A QTL with the Sr13 locus described by Simons et al. (2010). Significant associations were also detected in

other chr. regions not known to harbor stem rust resistance genes. In particular, associations were detected

on chrs. 1A, 3B, 4A, 5A and 5B. Notably, the two regions on chr. 5A represent novel loci for stem rust

resistance. In particular, the region associated to gwm291 showed significant effects in three out of four

seasons.

CONCLUSIONS AND PERSPECTIVES

Association mapping of the 183 durum wheat elite cultivars allowed us to:

i) identify significant genetic variation for stem rust response at the adult stage

ii) localize several putative QTLs involved in the stem rust response in fields in Ethiopia under high

infection rate

iii) confirm the role of the Sr13 locus previously described in biparental mapping populations

iv) highlight novel loci such as near gwm291 on chr. 5A with consistent effect on stem rust across seasons

Further work is in progress to verify the effectiveness of these novel loci in additional genetic backgrounds.

The molecular markers information herein reported can be exploited to pyramid the most beneficial

haplotype alleles to enhance stem rust resistance in durum wheat.

Molecular variation in the Durum Panel

Considering the SSRs markers, the level of polymorphism was quite high (5.1 allelic variants / SSR, on

average) with an average of 2.5 non-rare allelic variants / SSR (frequency ≥ 0.10). Such non-rare variants

were used to carry out the marker-trait association tests.

Acknowledgements

Table 2. Quantitative trait loci (QTLs) for stem rust response as identified through association mapping

in a panel of 183 elite durum wheat accessions evaluated in Debre Zeit, Ethiopia, with significant

effects observed over three to four evaluation seasons.

Descriptive statistics DZO-2009

(off

season)

DZM-2009

(main season)

DZO-

2010(off

season)

DZM-2010

(main season)

Mean over

seasons

Grand mean 41.7 33.2 48.7 28.9 38.1

Range 0-80 0-80 0-90 8-68 2.1-72.0

CV (%) - 26.6

LSD (P 0.05) - 20.9

Heritability - 0.80

Table 1. Descriptive statistics for field stem rust response (reported as Coefficient of Infection) in

the durum panel of 183 lines grown across four growing seasons.

The financial support of the Monsanto’s Beachell-Borlaug International Scholars Program

is gratefully acknowledged.

a: position of the QTL most associated marker as from the durum consensus map (Maccaferri et al.,

unpublished) used as reference b: range of percentage of the phenotypic variation (R2) for stem rust responses observed across the three to

four evaluation seasons with significant marker-trait association c: percentage of the phenotypic variation (R2) for stem rust responses averaged over the four evaluation

seasons that were accounted for by the QTL most associated marker

Chromosome Most

associated

marker

Position

(cM) a

R2 range

(%) b

R2

(%) c

Associated markers in the

QTL region

Interval

width(cM) a

1BS barc8 32.0 3.2 - 5.6 4.6 gwm1100, wPt-2999, wPt-

4605, tPt-8831, wPt-9864, wPt

-3582, wPt-4133, wPt-1876,

wPt-5899, wPt-4729

9.0

2AS gwm1045 87.7 3.3 - 5.8 3.9 gwm425, cfa2263 12.5

2BL wmc356 220.0 3.2 - 6.6 4.1 - 0.0

3AS wPt-7992 8.0 1.7 - 4.7 3.3 wPt-6854, barc12, wPt-1111 3.5

3AL wmc388 85.6 1.9 -4.1 4.0 - 0.0

5AL gwm126 93.3 1.8 -4.8 4.1 - 0.0

5AL gwm291 111.7 2.7 -5.7 4.4 - 0.0

6AL gwm427 139.5 1.7 -6.8 3.5 wmc580 0.1

6AL CD926040 144.0 3.5 - 11.3 7.1 wPt-9474, wPt-4229, wPt-

5654, wPt-3247,wPt-4663

9.3

6AL barc104 155.3 6.1 - 9.7 4.5 - 0.0

7AS wPt-2799 38.2 1.7 - 4.9 5.2 barc70, gwm1187, wmc479 6.3

7AS wPt-7785 94.8 1.1 - 2.3 1.5 - 0.0