Fig 1.
Stages of early panicle development in rice.
The rice panicle (A) has a defined architecture derived from the meristematic structure established early in inflorescence development (B, C), characterized by diagnostic changes in the gene expression pattern (D). (A) Panicle architecture. Representative primary branch (PB), secondary branch (SB), lateral and terminal spikelet (LS and TS) are labeled. (B) Developmental stages of the reproductive meristem, which produces primary branch meristems (PBM), secondary branch meristems (SBM), lateral and terminal spikelet meristems (LSM, TSM), which then give rise to floral meristems (LFM, TFM). (C) Light microscope images (scale bar = 100μm). (D) Diagnostic changes in gene expression occurring during early panicle development based on qRT-PCR using three biological replicates with reproductive stages as defined in (C). Statistical analysis performed by ANOVA with Holm post-test to compare stage 0 to stages 1–5 (* p < 0.05, ** p < 0.01).
Fig 2.
Differing expression levels of cytokinin-related genes in the early rice panicle based on NanoString analysis.
(A) Genes involved in cytokinin metabolism and signaling. (B, C) Box plot analysis for expression of genes involved in cytokinin metabolism (B) and signaling (C) based on NanoString analysis. Expression was analyzed for stages 0–5 of early panicle development. The bottom and top of each box indicate the 25th and 75th percentile for data expression values, the band in the middle of the box indicates the median expression value, and the ends of the whiskers indicate the minimum and maximum values. Expression values for the box-plot analysis are log2 transformed. All probes recognize single genes, except RR9 which also recognizes RR10, and RR13 which also recognizes RR8 and RR12.
Fig 3.
Differential expression of genes involved in cytokinin action between panicles and vegetative tissues.
The average gene expression value during early panicle development was compared to that found in rice roots and shoots following treatment for 2 h with 5 μM BA or a vehicle control [13]. Representative genes exhibiting significant differential expression in panicles compared to vegetative tissues are shown (ANOVA with Holm post-test, * p < 0.05, ** p < 0.01).
Fig 4.
Dynamic changes for expression of genes involved in cytokinin action during early panicle development.
A cluster analysis was performed based on Euclidian distance between gene expression at stages 0–5 of early panicle development. This is plotted as a heat map with a dendrogram. Genes are color-coded based on whether they are involved in cytokinin metabolism (in blue) or signaling (in orange), with differing shades for each of the families involved in these processes. Genes fall into three major subfamilies (I, II, and III) based on the cluster analysis.
Fig 5.
Confirmation of expression changes for representative genes during panicle development.
NanoString expression data is compared to expression data derived from RNA-Seq (A) or qRT-PCR (B). Genes are from subfamilies I (CKX3, HK6), II (CKX11, LOG, RR11), and III (RR24), based on the cluster analysis (Fig 4). NanoString expression is given in normalized nCounts. The RNA-Seq dataset from panicle meristems [29] covers stages 1 through 4 of early development and is given in transcripts per million (TPM). qRT-PCR data is given as a relative gene expression (RQ) based on an actin control gene.