The transcriptional landscape of polyploid wheat

TLDR: This study leverages 850 wheat RNA-sequencing samples, alongside the annotated genome, to determine the similarities and differences between homoeolog expression across a range of tissues, developmental stages, and cultivars and suggests that the transposable elements in promoters relate more closely to the variation in the relative expression of homoeologicals across tissues than to a ubiquitous effect across all tissues.

Abstract: The coordinated expression of highly related homoeologous genes in polyploid species underlies the phenotypes of many of the world's major crops. Here we combine extensive gene expression datasets to produce a comprehensive, genome-wide analysis of homoeolog expression patterns in hexaploid bread wheat. Bias in homoeolog expression varies between tissues, with ~30% of wheat homoeologs showing nonbalanced expression. We found expression asymmetries along wheat chromosomes, with homoeologs showing the largest inter-tissue, inter-cultivar, and coding sequence variation, most often located in high-recombination distal ends of chromosomes. These transcriptionally dynamic genes potentially represent the first steps toward neo- or subfunctionalization of wheat homoeologs. Coexpression networks reveal extensive coordination of homoeologs throughout development and, alongside a detailed expression atlas, provide a framework to target candidate genes underpinning agronomic traits in wheat.

Full text

University of Birmingham
The transcriptional landscape of polyploid wheat
Borrill, P; International Wheat Genome Sequencing Consortium (IWGSC)
DOI:
10.1126/science.aar6089
License:
None: All rights reserved
Document Version
Peer reviewed version
Citation for published version (Harvard):
Borrill, P & International Wheat Genome Sequencing Consortium (IWGSC) 2018, 'The transcriptional landscape
of polyploid wheat', Science, vol. 361, no. 6403, eaar6089. https://doi.org/10.1126/science.aar6089
Link to publication on Research at Birmingham portal
Publisher Rights Statement:
Checked for eligibility: 26/09/2018
This is the author’s version of the work. It is posted here by permission of the AAAS for personal use, not for redistribution. The definitive
version was published in Science on 17 August 2018, Volume 361, DOI: 10.1126/science.aar6089
General rights
Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the
copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes
permitted by law.
•Users may freely distribute the URL that is used to identify this publication.
•Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private
study or non-commercial research.
•User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?)
•Users may not further distribute the material nor use it for the purposes of commercial gain.
Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document.
When citing, please reference the published version.
Take down policy
While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been
uploaded in error or has been deemed to be commercially or otherwise sensitive.
If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access to
the work immediately and investigate.
Download date: 10. Aug. 2022

The transcriptional landscape of polyploid wheat
R. H. Ramírez-González
1†
, P. Borrill
1†*
, D. Lang
2
, S. A. Harrington
1
, J. Brinton
1
, L. Venturini
3
,
M. Davey
4
, J. Jacobs
4
, F. van Ex
4
, A. Pasha
5
, Y. Khedikar
6
, S. J. Robinson
6
, A. T. Cory
7
, T.
Florio
1
, L. Concia
8
, C. Juery
9
, H. Schoonbeek
1
, B. Steuernagel
1
, D. Xiang
10
, C. J. Ridout
1
, B.
Chalhoub
11
, K. F. X. Mayer
2,12
, M. Benhamed
8
, D. Latrasse
8
, A. Bendahmane
8
, International
Wheat Genome Sequencing Consortium
13
, B. B. H. Wulff
1
, R. Appels
14
, V. Tiwari
15
, R. Datla
10
,
F. Choulet
9
, C. J. Pozniak
7
, N. J. Provart
5
, A. G. Sharpe
16
, E. Paux
9
, M. Spannagl
2
, A.
Bräutigam
17‡
, C. Uauy
1*
Affiliations:
1
John Innes Centre, Norwich Research Park, NR4 7UH, Norwich, United Kingdom.
2
Plant Genome and Systems Biology, Helmholtz Center Munich, Ingolstaedter Landstr. 1 85764
Neuherberg, Germany.
3
Earlham Institute, Norwich Research Park, NR4 7UZ, Norwich, United Kingdom.
4
Bayer Crop Science, Innovation Center, Technologiepark 38, 9052, Zwijnaarde, Belgium.
5
Cell & Systems Biology/ Centre for the Analysis of Genome Evolution and Function,
University of Toronto, 25 Willcocks St., Toronto, ON. M5S 3B2, Canada.
6
Saskatoon Research and Development Centre, Agriculture and Agri-Food Canada, 107 Science
Place, Saskatoon, SK, S7N 0X2, Canada.
7
Crop Development Centre, University of Saskatchewan, Agriculture Building, 51 Campus
Drive, Saskatoon SK, S7N 5A8, Canada.
8
Institut of Plant Sciences Paris-Saclay (IPS2), UMR 9213/UMR1403, CNRS, INRA, Université
Paris-Sud, Université d’Evry, Université Paris-Diderot, Sorbonne Paris-Cité, Bâtiment 630,
91405 Orsay, France
9
GDEC, INRA, UCA, 5 Chemin de Beaulieu, Clermont-Ferrand 63039, France.
10
Aquatic and Crop Resource Development, National Research Council Canada, 110
Gymnasium Place, Saskatoon, Saskatoon S7N 0W9, Canada.
11
INRA, 2 rue Gaston Crémieux, Evry 9057, France
12
School of Life Sciences Weihenstephan, Technical University Munich, Germany
13
IWGSC, 5207 Wyoming Road, Bethesda, MD 20816, USA
14
BioSciences, University of Melbourne/AgriBio-LaTrobe University/Veterinary and Ag
Science, Murdoch University 90 South St, Perth Western Australia 6150, Australia.
15
Plant Sciences and Landscape Architecture, University of Maryland, 4291 Field House Drive,
College Park, Maryland, 20742 USA
16
Global Institute for Food Security, University of Saskatchewan, 110 Gymnasium Place
Saskatoon SK. S7N 4J8, Canada.
17
Molecular Genetics, IPK Gatersleben, Corrensstrasse 3, 06466 Gatersleben, Germany.
*Correspondence to: Cristobal Uauy (cristobal.uauy@jic.ac.uk) and
Philippa Borrill (philippa.borrill@jic.ac.uk)

†
These authors contributed equally to the work.
‡
Current address: Institute for Computational Biology, Faculty of Biology, University of
Bielefeld, 33501, Bielefeld, Germany.
Abstract:
The coordinated expression of highly related homoeologous genes in polyploid species underlies
the phenotypes of many of the world’s major crops. Here we combine extensive gene expression
datasets to produce a comprehensive, genome-wide analysis of homoeolog expression patterns in
hexaploid bread wheat. Bias in homoeolog expression varied between tissues, with ~30% of
wheat homoeologs showing non-balanced expression. We found expression asymmetries along
wheat chromosomes, with homoeologs showing the largest inter-tissue, inter-cultivar, and coding
sequence variation most often located in high-recombination distal ends of chromosomes. These
transcriptionally dynamic genes potentially represent the first steps towards neo/sub-
functionalization of wheat homoeologs. Co-expression networks revealed extensive coordination
of homoeologs throughout development and, alongside a detailed expression atlas, provide a
framework to target candidate genes underpinning agronomic traits in wheat.
One Sentence Summary: Homoeolog expression profiling across tissues reveals expression
asymmetry along wheat chromosomes.

Main Text:
Polyploidy arises from whole-genome duplication or interspecific hybridization and is ubiquitous
in eukaryotic plant and fungal lineages. Polyploidy has been proposed to confer adaptive
plasticity, thereby shaping the evolution of plants, fungi and, to a lesser degree, animals (1, 2).
This plasticity has facilitated the domestication and adaptation of several major crop species (3)
including hexaploid wheat (Triticum aestivum; AABBDD genome), which is derived from
relatively recent interspecific hybridizations between three different diploid species. In such
polyploids, gene duplication alters the transcriptional landscape (4) by providing additional
flexibility to adapt and evolve new patterns of gene expression for homoeologous gene copies
(5). This flexibility has been suggested to be an important mechanism for controlling adaptive
traits (6, 7), for example through neo-functionalization of duplicated genes (8) or tissue-specific
expression (9). However, despite the likely importance of polyploidy in affecting gene
expression, we have a limited understanding of the extent to which homoeologs share or differ in
their expression patterns, the spatiotemporal dynamics of these relationships, and how epistatic
interactions between individual homoeologs impact on biological traits. The new genomic
resources available for wheat (10), along with its meiotic stability (11) and syntenic gene order
(12), make wheat a particularly informative system for gaining insight into the effects of recent
polyploidy on gene expression.
In this study, we leverage available RNA-Seq data (529 samples from 28 studies) and add 321
samples to explore global gene expression in hexaploid wheat across a diverse range of tissues,
developmental stages, cultivars, and environmental conditions (13). We organized these sets of
RNA-Seq samples into partially overlapping datasets from (i) a single developmental time course
experiment (n = 209 samples), (ii) the reference accession Chinese Spring (CS) under non-stress
conditions (n = 123 samples), (iii) four main tissue types under non-stress conditions (n = 537
samples), and (iv) seedling samples from abiotic (n = 50) and biotic (n = 163) stress experiments
including controls (Table S1). These datasets, alongside a complete and annotated genome and
transcriptome (10), provide an opportunity to conduct homoeolog-specific transcriptome
profiling and to generate gene regulatory networks to better understand the spatiotemporal
coordination of individual homoeologs underlying trait biology on a genome-wide scale.

A developmental gene expression atlas in polyploid wheat
We first assessed expression patterns through a developmental time course of the commercial
wheat cultivar ‘Azhurnaya’ including 209 RNA-Seq samples comprising 22 tissue types
representing grain, root, leaf, and spike samples across multiple timepoints (Fig. 1). We
quantified expression using pseudoalignment of RNA-Seq reads to the RefSeqv1.0 transcriptome
as implemented in kallisto (14), which accurately quantifies reads in a homoeolog-specific
manner in polyploid wheat ((13, 15); Figs. S1-S2). We found evidence of expression for 83,741
(75.6%) high-confidence genes based on expression of >0.5 transcripts per million (TPM) in at
least one of the 22 tissue types, and conducted complexity (Table S2) and differential expression
analyses (Fig. S3). Tissue distinguished samples across development (Fig. S4; (13)) consistent
with observations in other plant and animal species (16, 17). However, within similar tissue
types, genome of origin also influenced expression patterns, consistent with previous results in
wheat grain samples (18). This gene expression atlas provides a valuable resource for breeders
and researchers to query and analyze their genes of interest through www.wheat-expression.com
(15) and the Wheat eFP Browser (http://bar.utoronto.ca/efp_wheat/cgi-bin/efpWeb.cgi; Fig. S5
(19)).
Homoeolog expression patterns
In polyploid wheat, quantitative variation for many agronomic traits is modulated by genetic
interactions between multiple sets of homoeologs in the A, B, and D genomes (20). These
interactions range from buffering effects observed when gene homoeologs are functionally
redundant (21), to dominance effects where variation in a single homoeolog can lead to dominant
phenotypes (22). Understanding how these interactions influence gene expression will help
inform strategies to target and manipulate individual or multiple homoeologs to quantitatively
modulate trait responses for crop improvement (20).
To determine patterns of homoeolog expression, we analyzed 123 RNA-Seq samples
representing 15 tissues under non-stress conditions (Table S1) from CS. This was the same
accession used to generate the reference genome (10), thereby excluding cultivar-specific
polymorphisms from our analysis. We found evidence of expression for 82,567 (74.5%) high-
confidence genes, consistent with the cultivar Azhurnaya developmental time course. We

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "University of birmingham the transcriptional landscape of polyploid wheat" ?

The authors found expression asymmetries along wheat chromosomes, with homoeologs showing the largest inter-tissue, inter-cultivar, and coding sequence variation most often located in high-recombination distal ends of chromosomes. These transcriptionally dynamic genes potentially represent the first steps towards neo/subfunctionalization of wheat homoeologs. 

Q2. What tests were used to test for differences in TE density between categories?

Whitney tests with Benjamini-Hochberg adjusted P values were used to test for differences in TE density between categories across each window. 

Q3. How many tiles were used to summarize the methylation ratios?

The two flanking regions and the gene feature were each divided into 50 tiles (150 tiles in total) to summarize the observed methylation ratios. 

Q4. How was the count expression level of each gene normalized?

The count expression level of each gene was normalized using variance stabilizing transformation from DESeq2 (66) to eliminate differences in sequencing depth between studies. 

Q5. What was the soft power threshold for the abiotic and grain network?

For the abiotic and grain network the 0.9 threshold was not crossed until 15 and 20 respectively, which may be due to strong differences between samples within these datasets, therefore the soft power threshold was selected according to the number of samples, resulting in abiotic = 7 and grain = 

Q6. What was the FIMO tool used to predict TFBS?

The FIMO tool fromthe MEME suite (v 4.11.4 (63)) was used with a position weight matrix (PWM) obtained from plantPAN 2.0 (64) to predict TFBS based on previously identified sites across multiple plant species. 

Q7. How many ng of nuclear DNA was used to assess the conversion efficiency?

A total of 500 ng of nuclear DNA was spiked with 270 pg of Lambda DNA to assess the conversion efficiency obtained using the EZ DNA Methylation-GoldTM Kit (Zymo research corp, Irvine, Ca, USA). 

Q8. How many modules were used to calculate the topographical overlap matrices?

The topographical overlap matrices (TOM) were calculated by the blockwiseModules function using TOMType = “unsigned” and the minimum module size was set to 30. 

Q9. How many balanced triads were found in each of the 15 tissues?

The authors found that 83.6% of balanced triads remained balanced in each of the 15 individual tissues,whereas dominant and suppressed triads tended to be more variable across tissues with only73.4% and 62.2%, respectively, staying within their global dominance group across all 15 tissues(Fig. 3A). 

Q10. How did the authors calculate the read density of each histone mark?

The authors calculated meta-gene profiles for each category by computing the read density of each histone mark over different triads categories using Deeptools (61) computeMatrix scale-regions and plotted it with plotProfile. 

Q11. What is the significance of the co-expression networks?

To understand how this coordination ofhomoeolog spatiotemporal expression may influence biological processes, the authors developed a seriesof co-expression networks to provide insight into tissue-specific developmental and stress-relatedprocesses. 

Q12. How did the authors calculate the distance between the homoeologs in each triad?

The authors calculated the Euclidean distance between module eigengenes using the R package dist() and with these values the authors calculate the distances between the homoeologs in each triad. 

Q13. What is the role of TBF1 in the transition from growth to defense?

The Arabidopsis ortholog of these genes, TBF1, wasoriginally identified for its role in pathogen defense response (42), and has been shown to play akey role in the transition from growth to defense (43) while also positively regulating acquiredthermotolerance (44). 

Q14. What was the soft power threshold for each network?

The soft power threshold was calculated as the first power to exceed a scale-free topology fit index of 0.9 for each network separately. 

Q15. What is the significance of the CG methylation mark in angiosperms?

Gene body CG methylation is widely conserved in angiosperms, although its functionalsignificance is currently under debate (28, 29) given that two angiosperm species lack thisepigenetic mark altogether (30).