Genetic dissection of drought tolerance in potato

Drought is the most important cause of crop and yield loss around the world. Breeding for

drought tolerance is not straightforward, as drought is a complex trait. A better understanding

of the expression of drought traits, the genes underlying the traits and the way these genes

interact will significantly increase the success of breeding for drought tolerance.

Potato is an important food crop, yet it is relatively susceptible to drought. As a first step

towards identifying the genetic basis for drought tolerance in potato, we make use of diploid

potato populations that have been genetically well characterized (CxE, SHxRH). The CxE

population was extensively evaluated for drought tolerance in vitro and for two successive

years (2008, 2009) under greenhouse conditions and the data were used for QTL mapping.

For optimal QTL mapping, we expanded the CxE and SHxRH genetic maps with 499 SNP

markers (two arrays 384 and 768SNP arrays respectively, enriched for putative stress

tolerance candidate genes). The SNPs were discovered in public EST databases using

QualitySNP software and detected with the Illumina GoldenGate assay. About 300 SNPs

served as bridge markers between the CxE and SHxRH maps. This will enable us to make use

of the extensive genetic and sequence information of the SHxRH population and the RH

genome sequence. With the availability of the potato genome sequence of the doubled

monoploid DM1-3 516R44 (DM) (www.potatogenome.net), it was possible to further

examine the SNP marker loci for paralogs and intron spanning sequences. In total 732 SNP

marker loci were found to be unique in the potato genome sequence. Many of these SNP

markers not only served as landmarks on the genetic map but may also as putative genes

underlying quantitative traits. In addition the validated SNP markers are now utilized as

anchors in the potato physical map.

We investigated the possibility of screening potato for relevant drought traits in in vitro

cultures and evaluated the CxE population for the response to PEG-induced water deficit

stress and recovery potential after stress. Significant genetic variation was observed for the

response to drought and for recovery potential. Several shoot and root growth traits were

measured. In this study the genetic variation and heritability estimates were high to very high

for the measured traits under control and recovery condition. In total 23 QTLs were detected

in plants under control, stress and recovery treatments. Interesting putative candidate genes

that may underly stress response QTLs were identified.

The drought tolerance evaluation of the CxE population in pots in the greenhouse included

traits like leaf Relative Water Content, δ13C as a measure of Water Use Efficiency,

Chlorophyll Fluorescence, Chlorophyll Content, shoot and root biomass and tuber yield. The

progeny displayed a wide contrast for drought tolerance, with individuals surviving and

recovering completely after 3 weeks of drought, and others completely wilted beyond

recovery. Most of the traits had high heritabilities. QTLs effective in multiple treatments and

years were detected for tuber number, tuber weight, plant height, shoot fresh and dry weight.

Other QTLs were found to be dependent on the environment: QTL x Environment interaction

was found for leaf d13C under drought conditions and we speculate that the function of δ13C

was genetically split into a stomatal and non-stomatal component.

Many of the QTLs for growth traits measured both in the greenhouse and in in vitro cultures

were specific to either of the growth conditions. Yet significant QTLs that were detected for

plant height, shoot dry weight, fresh biomass for plants grown in the greenhouse were also

found when the population was grown in vitro. These QTLs may be less affected by

environmental influences, and we may therefore expect that some of these QTLs will be

relevant under field conditions as well. This also suggests that the in vitro system may be used

for preliminary selection in breeding programmes for specific performance-related traits.

The genetic architecture of transcript-level variation for drought response was captured in the

potato population CxE and mapped as expression QTLs (eQTLs). We anchored the

differentially expressed genes to the genome sequence of potato, and this enabled us to

determine whether the transcription of these genes (the eQTLs) is in cis or in trans regulated.

The combined use of genome-wide detection of eQTLs in combination with genome sequence

information for gene location has enables us to detect regulatory hot spots for drought

response in the CxE population. Based on gene ontology annotation, a number of eQTLs were

detected for genes known to be involved in drought signal transduction and drought-induced

transcriptional regulation, and for redox genes. Examination of co-localization of eQTLs and

phenotypic QTLs identified several interesting eQTLs for genes that may be involved in

specifying the phenotypic QTL, for instance, the eQTL for a gene that was annotated with a

putative function in the photosystem II light reaction colocalized with trait QTL of

chlorophyll florescence (Fv/Fm) on chromosome 1, along with other genes involved in 139

drought response such as heat shock proteins and signaling proteins with known induced

expression under stress conditions. On chromosome 10, eQTLs for genes involved in carbon

partitioning, signaling receptor kinases, transcription factors and hormone and lipid

metabolism were colocalized with phenotypic QTLs for chlorophyll content and stomatal

component of δ13C. As we have only touched the surface of the information contained in the

transcriptome dataset combined with the phenotyping data, continued efforts on mining the

dataset and in depth analysis will most likely reveal more putative candidate genes for QTL

effects.

This thesis constitutes the first knowledge of in vitro and greenhouse screening for drought

tolerance in potato and has led to the description of important traits for screening and

selection in breeding for drought tolerance. The QTLs identified in this thesis may be

interesting targets for potato breeding to improve drought tolerance of the potato crop.

Furthermore, our results illustrate the power of application of integrated genetic and genomics

approaches to unravel the molecular components underlying abiotic stress tolerance traits.