For the
L. erythrorhizon root periderm and vascular tissues RNA-seq experiment, 3-month-old Siebold & Zucc. plants grown in soil under standard greenhouse conditions were harvested. Roots were collected from nine individual plants and divided into three groups, each containing three unique individuals. The periderm and vascular tissues were isolated by peeling the periderm from the roots (Fig. S
5a), and the prepared portions from the three individuals in each group were pooled. Tissues were frozen in liquid nitrogen, ground by mortar and pestle, and 100 mg was used to analyze total shikonin content each sample (Fig. S
5b). From the same sets of samples, RNA was extracted as described below, quantified, and DNase-treated (NEB) according to the manufacturer’s instructions. A total of six cDNA libraries from the three biological replicates prepared from each of the
L. erythrorhizon periderm and vascular tissue pools, were constructed using a ribominus TruSeq Stranded Total RNA library prep kit (Illumina, San Diego, CA), and 101-bp paired-end reads were generated via Illumina HiSeq 2500 at the Purdue Genomics Center, with at least 67 million reads per library. Sequence quality was assessed by FastQC (v. 0.10.0;
http://www.bioinformatics.babraham.ac.uk). The raw data were submitted to the Sequence Read Archive (
http://www.ncbi.nlm.nih.gov/sra/) and are available at the NCBI Sequence Read Archive (
PRJNA596998).
The experimental design for the RNA-seq experiment comparing
L. erythrorhizon hairy roots sampled in B5 in the light and M9 in the dark was based on a previous report of observed rapid increases in expression of shikonin precursor pathway genes, and in
PGT, within 2 h after switching
L. erythrorhizon cell cultures from growth in B5 in the light to growth in M9 in darkness
40 (link). In this study, several cultures from three independently generated
L. erythrorhizon hairy root lines were started in liquid Gamborg B5 media containing 3% sucrose at 28 °C in the light (~100 µE m
−2 s
−1). After 2 weeks, hairy roots from three cultures for each of the three lines (
n = 3 biological replicates per line) were harvested and pooled to represent the B5 light-treated samples. The remaining hairy root cultures were transferred to M9 media and darkness. After 2 h, hairy roots from three cultures for each of the three lines (
n = 3 biological replicates per line) were harvested and pooled to represent the M9 dark-treated samples. Samples were frozen in liquid nitrogen, ground by mortar and pestle, and RNA was extracted as described below. Six cDNA libraries were generated with a TruSeq Stranded mRNA library prep kit (Illumina, San Diego, CA) and were sequenced on an Illumina NovaSeq 6000 at the Purdue Genomics Center. Sequence quality assessment were performed as described above for the periderm and vascular tissues RNA-seq experiment. The raw data were submitted to the Sequence Read Archive (
http://www.ncbi.nlm.nih.gov/sra/) and are available at the NCBI Sequence Read Archive (PRJNA596998).
Additionally, unstranded RNA-seq data of
L. erythrorhizon whole roots and aerial tissue from an unknown accession was downloaded from the NCBI SRA (experiments SRR3957230 and SRR3957231) to include in the gene expression analysis. Gene abundance estimates of PGT and PGT-like genes (Fig.
3b, Table S
9) were measured using Kallisto v0.45.0
69 (link) and normalized for library depth using DESeq2
70 (link). Differential expression status was determined using the EdgeR v3.24.3
71 (link) package. For the EdgeR analysis, raw counts were normalized into effective library sizes using the trimmed mean of M-values (TMM) method
72 (link), and exact tests were conducted using a trended dispersion value and a double tail reject region. A false discovery rate was calculated using the Benjamini–Hochberg procedure
73 . Genes with a log
2-fold change in abundance greater than 1 and false discovery rate less than 0.05 were considered as differentially represented.