Article Figures & Data
Figures
- Figure 1.
Hierarchical Clustering of Samples and Genes across Developing Xylem and Phloem Tissues.
(A) Transverse cross-section image from one of the sampled trees (tree T1). The pooled samples used for RNA-seq are visualized by overlaying them on the transverse cross section (positions of the pooled samples on the section are approximated). The color bar below the image shows four sample clusters identified by hierarchical clustering (see [B]). The color bar above the image shows the estimated tissue composition for each sample.
(B) Hierarchical clustering of all 106 samples from the four replicate trees using mRNA expression values for all expressed genes. The four main clusters are indicated with colors.
(C) Heat map describing hierarchical clustering of the 28,294 expressed annotated genes using mRNA expression values for all samples. Expression values are scaled per gene so that expression values above the gene average are shown in red and below average in blue. Eight main clusters have been assigned colors and are denoted a to h.
(D) Average expression profiles in tree T1 for each gene expression cluster and distinct subclusters (solid white lines). The expression profiles of all individual genes assigned to each cluster are shown as gray lines in the background.
- Figure 2.
The Expression Profiles Are Highly Reproducible across Trees.
(A) Hierarchical clustering of genes across developing xylem and phloem tissues with samples ordered according to sampling order for each of the four replicate trees. The color bars indicate sample and gene clusters (see Figure 1).
(B) Expression profiles of the secondary cell wall CesA genes in all four trees (T1–T4). PtCesA4 (Potri.002G257900), PtCesA7-A (Potri.006G181900), PtCesA7-B (Potri.018G103900), PtCesA8-A (Potri.011G069600), and PtCesA8-B (Potri.004G059600).
- Figure 3.
A Modular Version of the Coexpression Network.
Genes with representative expression profiles were identified in the coexpression network (at a Z-score threshold of 5) by iteratively selecting the gene with the highest centrality and a coexpression neighborhood not overlapping with any previously selected genes’ neighborhood. Only annotated genes and positively correlated coexpression links were considered (i.e., Pearson correlation > 0). The selected genes and their coexpression neighborhoods (network modules) were represented as nodes in a module network. The modules were numbered according to the order in which they were selected (and hence according to size) and given descriptive names based on gene function enrichment analysis (Supplemental Data Set 8). The nodes are colored according to the hierarchical clusters in Figure 1 and reflect the proportion of genes in each network module belonging to the different hierarchical clusters. Nodes were linked if the neighborhoods overlapped at a lower coexpression threshold (Z-score threshold of 4). Link strengths are proportional to the number of common genes. Overlaps of fewer than five genes were not represented by links, and only the 41 network modules with at least 20 genes were displayed. Asterisks next to the module names indicate conservation in Norway spruce: *, 6–24% of the genes have conserved coexpression neighborhoods; **, 25–50%; and ***, >50% (Supplemental Data Set 8D).
- Figure 4.
Similarities in Expression Patterns for Regulators of Early Vascular Differentiation between Primary and Secondary Growth.
(A) Expression profiles for the homologs of OPS, BRX, CLE45, and BAM3. PtOPS-A was highly expressed in the cambium, while PtBRX-A, PtCLE45-A, and PtBAM3-A were highly expressed in the secondary phloem.
(B) Expression profiles for the homologs of APL and NAC45/86. PtAPL and PtNAC genes were highly expressed in the secondary phloem.
(C) Regulators of periclinal cell divisions. Expression profiles for genes homologous to MP, TMO5, LHW, and LOG4 genes. High expression levels for all genes overlap in the expanding xylem. For the complete list of genes, see Supplemental Data Set 9.
- Figure 5.
Primary and Secondary Cell Wall Biosynthesis Genes.
(A) Expression profiles for pectin and xyloglucan biosynthetic genes. All genes are highly expressed during primary wall biosynthesis, but also at later stages during xylem development. (i) Representative expression for homogalacturonan biosynthesis genes (illustrated by PtGAUT1-A); (ii) expression pattern of PtGALS1; (iii) representative pattern of other RG-I biosynthesis genes (represented by PtGALS2-A); and (iv) representative expression pattern for three of the putative xyloglucan biosynthesis genes (illustrated by PtXXT1-B). For a list of identified putative pectin and xyloglucan biosynthesis genes, see Supplemental Data Set 9.
(B) Expression patterns for CesA genes. (i) Members responsible for cellulose biosynthesis in the secondary wall layers are all induced in SCW biosynthesis zones in the xylem and phloem (illustrated by PtCesA4); (ii) members classified as primary wall CesAs typically peak in primary wall biosynthesis zone, but are also highly expressed during later stages of xylem differentiation (illustrated by PtCesA6-A); and (iii and iv) some members even peak during these later stages (illustrated by PtCesA6-C and PtCesA6-F). For a complete list of putative CesA genes, see Supplemental Data Set 9.
(C) The GT43 gene family responsible for xylan biosynthesis comprises three clades, A/B, C/D, and E, each having different expression here illustrated by (i) PtGT43A, (ii) PtGT43C, and (iii) PtGT43E. Expression profiles support the hypothesis that PtGT43A/B and PtGT43C/D are members of secondary wall xylan synthase complex, whereas PtGT43E and PtGT43C/D are members of the primary wall xylan synthase complex. For a complete list of genes coregulated with the three clades of GT43 genes, see Supplemental Data Set 9.
- Figure 6.
Expression Profiles of Genes Involved in the Lignification of Xylem Cells.
(A) Hierarchal clustering and heat map of the 59 monolignol biosynthesis genes expressed in AspWood. Expression values are scaled per gene so that expression values above the gene average are represented by red and below average by blue.
(B) Expression profiles of C4H (Potri.013G157900), C3H (Potri.006G033300), and F5H (Potri.005G117500) homologs.
(C) and (D) Hierarchical clustering of the 34 expressed laccase phenoloxidases (C) and expression profiles of representative genes (D): homologs of LAC4 (Potri.001G248700 and Potri.016G112100), LAC11 (Potri.004G156400), and LAC17 (Potri.001G184300, Potri.001G401300, and Potri.006G087100).
(E) and (F) Hierarchical clustering of the 42 expressed peroxidase phenoloxidases (E) and expression profiles of representative genes (F): homologs of PXR3 (Potri.003G214800), PXR25 (Potri.006G069600), PXR56 (Potri.007G122200), and PXR72 (Potri.005G118700 and Potri.007G019300).
- Figure 7.
NAC Domain Transcription Factors Are Expressed in Distinct Patterns Corresponding to Phylogenetic Clustering.
(A) Phylogenetic tree of Populus wood associated NAC-domain transcription factors with homologs in Arabidopsis. Colors indicate coexpression. Arabidopsis gene names in bold are used as clade names.
(B) Expression profiles of Populus SND2 homologs (Potri.007G135300, Potri.017G016700, Potri.004G049300, and Potri.011G058400) compared with the VND3 homolog with a similar expression profile (Potri.007G014400).
(C) Expression profiles of Populus SND1 homologs (Potri.001G448400, Potri.002G178700, Potri.011G153300, and Potri.014G104800).
(D) Expression profile of the Populus VND7 homolog expressed in AspWood (Potri.013G113100; Potri.019G083600 is not expressed).
(E) Expression profiles of Populus VND6 homologs (Potri.001G120000, Potri.003G113000, Potri.012G126500, and Potri.015G127400) compared with the VND3 homolog with a similar expression profile (Potri.005G116800).
- Figure 8.
WGD Paralog Expression.
(A) Circos plot of regulatory diverged and conserved paralogs. The clusters are ordered according to their peak of expression along the wood developmental gradient. An additional cluster for genes not expressed (ne) in our data is also added. Each cluster occupies a share of the circle’s circumference proportional to the number of genes in that cluster belonging to a paralog pair. Paralogs expressed in two different clusters are shown by links. The width of a link is proportional to the number of paralogs shared between these clusters (i.e., diverged pairs). Only pairs in different clusters and with an expression Pearson correlation < 0.5 were considered diverged. Links representing more pairs than expected by chance (P < 0.0001) are colored in a darker tone. The portion of a cluster without any outgoing links to other clusters represents the proportion of paralogs in this cluster where both genes belong to the cluster (i.e., conserved paralog pairs).
(B) Example of paralogs with highly diverged profiles (correlation = −0.78): FAAH (fatty acid amide hydrolase, Potri.005G070300 [cluster e] and Potri.007G098600 [cluster g]).
(C) Previously published real-time PCR data of CesA genes showed that PtCesA8-B (Potri.004G059600) was expressed higher than PtCesA8-A (Potri.011G069600) in secondary phloem and xylem (Takata and Taniguchi, 2015). Although these genes have highly similar expression profiles (Pearson correlation = 0.93) and are therefore considered as conserved according to our definition (both belong to cluster g), the data confirm the difference in expression levels and show that even more subtle regulatory divergence can be detected in the data set.
(D) The distribution of different gene sets among expression clusters: blue, all 28,294 expressed, annotated genes in our data; green, all 3729 paralog pairs with conserved expression (i.e., in the same expression cluster); red, 721 genes with high expression (among the top 5% most highly expressed genes in our data; 1413 genes) and in paralog pairs with conserved expression.
Additional Files
Supplemental Data
Files in this Data Supplement:
- Supplemental Figures and Tables
- Supplemental Dataset 1
- Supplemental Dataset 2
- Supplemental Dataset 3
- Supplemental Dataset 4
- Supplemental Dataset 5
- Supplemental Dataset 6
- Supplemental Dataset 7
- Supplemental Dataset 8
- Supplemental Dataset 9
- Supplemental Dataset 10
- Supplemental Dataset 11
- Supplemental Dataset 12
