NRI s mission is to provide distinctive high quality

“NRI's mission is to provide distinctive, high quality and relevant research, consultancy, teaching and advice in support of sustainable development, economic growth and poverty reduction. ” Next generation gene mining to decipher CBSV resistance in cassava Hale Ann Tufan Natural Resources Institute University of Greenwich

Outline • Introduction • Material and methods • General description of RNAseq data • RNA-seq data analysis • • • Clustering and expression profiles Gene ontology Genes of interest • Conclusions www. iita. org

Threat of CBSD • Genus Ipomovirus, family Potyviridae • Losses of US$ 100 million annually • Serious threat to cassava production in Eastern and Central Africa • Spread mechanically and by whitefly vector • Pressing need for new sources of resistance Herrera Campo et al. , (2011) Food Security, 3: 329 -345

Next generation sequencing for resistance gene discovery • For sequenced genomes, RNA-seq has potential to serve as a transcriptomics tool as well as marker development platform • Lower cost of sequencing enables use of this technology for resistance gene discovery Varshney et al. (2009). Trends in Biotechnology, 97: 522 -530

Approach Resistant and Susceptible lines Inoculate with virulent CBSV isolate Collect RNA from Control and CBSV infected plants Library construction and sequencing Data analysis Candidate genes Validation Test on cross progeny

Susceptible cv. Albert Leaves show severe symptoms and plants continue to show symptoms through development Roots show symptoms of rotting.

Resistant cv. Kaleso (Namikonga) Leaves show infection early but plant look and grow ‘normal’ thereafter Roots also show no sign of symptoms.

Methods • RNA isolated from 3 independent biological replicates each from 4 treatments: Albert Control, Albert CBSV, Kaleso Control, Kaleso CBSV • Pool replicates after quality control • RNA samples to GATC Biotech for sequencing • Illumina Hi. Seq 2000 platform, single-end 50 bp reads • Sequence reads mapped against reference genome with BWA aligner • The expression table buildup made by GATC in-house software

General description of data • ~50 million reads per sample, 50 -60% of reads mapped per sample • 34, 151 genes total Albert Control Albert CBSV Kaleso Control Kaleso CBSV Number of Reads Percentage All 54, 045, 667 - 60, 070, 579 - 38, 949, 010 - 49, 681, 907 - Mapping to whole genome 31, 632, 660 59 35, 964, 664 60 20, 946, 755 54 29, 534, 087 60 Non uniquely mapped 8. 674, 373 27 10, 282, 664 29 5, 526, 455 26 7, 563, 418 26 Uniquely mapped 23, 261, 749 74 26, 036, 303 72 15, 618, 148 75 22, 243, 065 75 Resulting Reads 23, 261, 749 74 26, 036, 303 72 15, 618, 148 75 22, 243, 065 75

General description of data • 28, 667 genes expressed in at least one of 4 treatments • Majority of these expressed in all treatments • High number of Kaleso-specific genes, compared to other treatments

Data analysis • Samples are pooled-limited options for data analysis • Genesis software used to analyze data http: //genome. tugraz. at/genesisclient_description. shtml • Co. V cutoff of 70% to identify genes with ‘significant’ gene induction between treatments Min Mean Max St. Dev % Co. V 3. 1 3. 32 3. 52 0. 17 5. 11 2. 7 3. 34 4. 37 0. 73 21. 96 0. 24 0. 39 0. 65 0. 19 47. 69 0. 009 0. 02 0. 04 0. 02 77. 92 0. 16 1. 63 3. 19 1. 65 101. 29 • K-mean clustering to identify groups of genes with similar expression patterns (50 iterations, specify 5 clusters)

K-Means Clusters Expression Profiles Cluster 1 133 Genes Kaleso CBSV specific (highly expressed) Cluster 3 150 Genes Albert specific Cluster 2 86 Genes Kaleso specific Cluster 4 4180 Genes Largely unchanged/ low expression (image truncated) Cluster 5 670 Genes Mix/ some tendency for higher expression in Kaleso CBSV

V o. C BS Ka les tro V CB S l tro Co n o. C on Ka les Alb ert l V o. C BS les Ka l tro V l tro CB S Co n o. C on les Ka ert Alb Gene Ontology Cluster 1 Cluster 2

V o. C BS Ka les tro V CB S l tro Co n o. C on Ka les Alb ert l V o. C BS les Ka l tro V l tro CB S Co n o. C on les Ka ert Alb Gene Ontology Cluster 3 Cluster 5

Genes highly upregulated in Kaleso CBSD (Cluster 1) • Metabolism: sucrose synthase, Fatty acid hydroxylase, hydroxycinnamoyl-Co. A shikimate/quinate hydroxycinnamoyl transferase • Transcription factors: MYB domain protein, zing finger domain protein, NAC transcription factor, WRKY protein • Signaling: MAPKK, MAPKKK, Leucine-rich repeat transmembrane protein kinase • Defence related: Seven transmembrane MLO family protein, peroxidase, pleiotropic drug resistance 1, Disease resistance-responsive (dirigent-like protein) family protein

Genes upregulated in Kaleso CBSD (Cluster 5) • Metabolism: Cinnamyl alcohol dehydrogenase 9 • Transcription factors: MYB domain protein, NAC domain protein, RWP-RK domain-containing protein, WRKY DNA-binding protein • Signaling: Protein kinase, receptor-like protein kinase 1, receptor serine/threonine kinase, Leucine-rich repeat protein kinase family protein, BAK 1 -interacting receptorlike kinase 1, cysteine-rich RLK (RECEPTOR-like protein kinase) • Defence related: disease resistance family protein, peroxidase, cellulose synthase, chitinase, beta glucosidase 11, Pathogenesis-related thaumatin, jasmonate -zim-domain protein 1, ethylene responsive element binding factor 4, ACC synthase 1, ethylene-responsive element binding factor 13, PR-1 • Other: RNA-dependent RNA polymerase 1, phloem protein 2 -B 15,

What’s next? Genes of interest Unique. ID cassava 4. 1_001246 m|PACid: 17989248 cassava 4. 1_025993 m|PACid: 17982084 cassava 4. 1_000944 m|PACid: 17992385 cassava 4. 1_021672 m|PACid: 17989194 cassava 4. 1_000627 m|PACid: 17974869 best arabidopsis TAIR 10 hit name best arabidopsis TAIR 10 hit symbol best arabidopsis TAIR 10 hit defline AT 3 G 07040. 1 RPM 1, RPS 3 NB-ARC domain-containing disease resistance protein AT 3 G 46710. 1 NB-ARC domain-containing disease resistance protein AT 4 G 12010. 1 Disease resistance protein (TIR-NBS-LRR class) family AT 1 G 61190. 1 LRR and NB-ARC domains-containing disease resistance protein AT 5 G 17680. 1 disease resistance protein (TIR-NBS-LRR class), putative

Model Gomez et al (2009) Eur. J. Plant. Path, 125: 1 -22 Modified from Maule et al. (2007) Mol. Plant Path. 8: 223– 231

Conclusions • Pooling samples yields good results for a snapshot study • Large number of genes specific to Kaleso CBSV treatment • Data analysis resulted in clusters of interesting genes, subset with large upregulation in response to Kaleso CBSV • Orthologues of genes well characterized to be involved in resistance responses are upregulated in response to Kaleso CBSV • Limitations in experimental design- focus on dominant resistance genes (NBS-LRR) for validation and further analysis • Knowledge can possibly be applied in the field- access to Albert x Kaleso cross progeny could yield very interesting results.

Thank you Please contact Dr. Maruthi Gowda at M. N. Maruthi@greenwich. ac. uk for further questions