Global dissection of cis and trans regulatory variations in Arabidopsis thaliana Xu Zhang

1. Global dissection of cis and trans regulatory variations in Arabidopsis thaliana Xu Zhang Borevitz Lab

2. Col and Van are collected from distinct geographic locations

3. Expression regulatory variation can act in cis or in trans cis polymorphism trans polymorphism gene 1 gene 2 gene 3 regulator 1 regulator 2

4. Dissection of cis and trans variation • eQTL mapping in segregating populations • CAUSAL cis and trans variation can be mapped individually • Additive effects and non-additive interactions among the eQTL can be assessed • limited detection power due to the sample size and extent of recombination of mapping population; it actually maps LOCAL and DISTANT variation

5. Allele specific expression in a heterozygous system allele A allele B Allele A and B are exposed to the same pool of trans factors. Thus allelic expression should be due to cis difference within or nearby the gene

6. Composite trans effects are inferred by comparison of hybrids and parents cis effect expression level Col Van hybrids cis effect trans effect expression level Col Van hybrids

7. Dissection of cis and trans variation • Allele specific expression in a heterozygous system • closely locates cis causal variation • composite effect of trans variation can be inferred by testing the deviation of allelic expression in heterozygous from the differential expression between homozygous parents • individual causal loci can NOT be mapped, their interaction can NOT be assessed

8. Biological samples for microarray hybridization Col ♀ x Van ♂ Col ♀ x Col ♂ Van ♀ x Van ♂ Van ♀ x Col ♂ • 4 replicates for parental gDNA1:1 mix • 4 replicates for parental mRNA1:1 mix • 4 replicates for Col (mother) x Van F1s • 4 replicates for Van (mother) x Col F1s

9. cis and trans The measurement: allele intensity ratio of the transcript – Col allele/Van allele • parental expression difference mRNA 1:1 mix vs gDNA 1:1 mix • cis regulatory difference F1s vs gDNA 1:1 mix • trans regulatory difference F1s vs mRNA 1:1 mix • imprinting F1c vs F1v

10. AtSNPtile: A SNP/tiling array • ~1.4M tiling probes at ~35bp resolution • ~1M SNP probes for 250K SNPs each SNP 2 alleles x 2 strands • allele A antisense strand: GACCAATTTTGACCCTAGATCGCCA • allele A sense strand : CTGGTTAAAACTGGGATCTAGCGGT • allele B antisense strand: GACCAATTTTGAACCTAGATCGCCA • allele B sense strand : CTGGTTAAAACTTGGATCTAGCGGT

11. Allele intensity ratio as a measurement of allelic difference log (A / B) of probe intensity log (A / B) of template mixture target amount in ug Across strands and SNPs A: Col allele; B: Van allele

12. The effect of overall target amount is small Linear regression for each SNP and strand

13. SNPs within transcribed region

14. The model

15. cis only Here cis variation up-regulates Van allele

16. trans only Here trans variation up-regulates Col allele

17. FLC reciprocal F1 hybrids Van Col FLC – cis and trans cis variation up-regulates Col allele trans variation up-regulates Van allele

18. FRIGIDA FRIGIDA – No difference

19. Summary for 12,311 analyzed genes An over-estimation of cis variation genes

20. The direction of cis and trans effects relative to that of parental expression difference

21. The size of cis and trans effects relative to that of parental expression difference cis vs parental difference trans vs parental difference trans vs cis

22. Validation for microarray result by single base extension coupled with Mass-spectrometry (ρ = 0.74, p < 1.08E -06, n=32)

23. Regional difference in sequence polymorphism between cis and trans genes

24. Gene regulatory network Regulatory genes Structural genes (Wittkopp 2004)

25. Chromosomal distributions of cis and trans genes Sliding window of 120 genes

26. Correlation of chromosomal distributions among cis, trans, sequence polymorphismand gene distance Sliding window of 120 genes

27. CG methylation for cis or trans genes CG methylation within promoter: repression gene expression CG methylation within gene proper: facilitate gene expression

28. Histone modification for cis and trans genes H3K27me3: Histone 3 lysine 27 trimethylation H3K9me3: Histone 3 lysine 9 trimethylation LND: low nucleosome density regions

29. Gene expression specificity for cis and trans genes expression level expression entropy gene length Data from Schmid et al, 2005 63 diverse tissues on Col wild type background

30. Conclusions • Large cis effect but more trans effect • cis genes tend to locate in polymorphic, gene-poor chromosomal regions, where repressive epigenetic modifications are enriched and gene expression is tightly regulated • trans genes tend to locate in conserved, gene-rich chromosomal regions, where activating epigenetic modifications are dominant and gene expression is more constitutive

31. Allele specific intron expression • Intron retention is common in plant • Allele specific intron expression suggests differential intron splicing

32. Summary for 6,707 analyzed introns

33. Distribution of sequence polymorphisms along up- and down-stream exons and the cis intron intron upstream exon downstream exon

34. Conclusions • extensive intron splicing variation • largely contribute by cis variation, no trans effect detected

35. Acknowledgements Borevitz Lab: Justin Borevitz Yan Li Christos Noutsos Geoffrey Morris Andrew Cal Paul Grabowski Traci Viinanen Whitney Panneton Greenhouse: Judy Coswell Sandra Suwanski John Zdenek