**4. Materials and Methods**

#### *4.1. Plant Materials and Culture Conditions*

*C. roseus* CMCs were established using the reported method [25], and *C. roseus* DDCs were established using the method of [29]. *C. roseus* CMCs were placed on pH-adjusted (5.75–5.80) Murashige and Skoog (MS) solid medium [30] supplemented with 2.0 mg/L α-naphthalene acetic acid (NAA), 10 g/L sucrose, and 4 g/L gellan gum, cultured at 25 ◦C in darkness, and maintained by serial subculturing every 12 days [25]. *C. roseus* DDCs were sub-cultured under the same condition as that of *C. roseus* CMCs.

Four or six days prior to the fungal elicitor experiments, CMCs were inoculated into 100 mL of MS liquid medium supplemented with 2.0 mg/L NAA and 20 g/L sucrose in a 250-mL Erlenmeyer flask at concentrations of 50.0 g fresh weight/L [25] and grown at 25 ◦C at 110 rpm with a 24-h light photoperiod. The suspension culture condition of *C. roseus* DDCs was the same as that of *C. roseus* CMCs.

#### *4.2. Procurement, Fungal Culture, and Preparation of Elicitor*

*A. flavus* elicitors were prepared using the methods of Tonk et al. [24] with slight modifications. The *A. flavus* strain (CGMCC No. 3.6434) was obtained from the China General Microbiological Culture Collection Center (CGMCC, Beijing, China). The fungus was grown at 28 ◦C in 100-mL Erlenmeyer flasks containing potato dextrose agar (BAM Media M127, Guangdong Huankai Microbial Sci.&Tech.Co., Ltd., Guangzhou, China) and maintained by serial subculture every 7 days. After 7 days of subculture, the mycelium was separated from the fungal mat with forceps.

The mycelium was washed several times with distilled water and homogenized with distilled water at 5.0, 15.0, and 25.0 g/L. Then, *A. flavus* mycelium elicitors were sterilized at 120 ◦C for 20 min and stored at 4 ◦C. At the same time, *A. flavus* medium elicitors were prepared from the fungal mat by the same method used to prepare *A. flavus* mycelium elicitors and stored at 4 ◦C.

#### *4.3. Induction Treatment with Fungal Elicitor*

The *A. flavus* fungal elicitor (0.1 mL) (*A. flavus* mycelium elicitor or *A. flavus* medium elicitor) was added to 4-day-old or 6-day-old suspensions of CMCs at final concentrations of 5, 15, and 25 mg/L. The check group (CK) was treated with 0.1 mL sterile distilled water. Then, the CMCs were further cultured at 25 ◦C at 110 rpm with a 24-h light photoperiod. After culturing for another 12, 24, 36, 48, and 60 h, the cultures and liquid media were harvested.

#### *4.4. Measurement of Dry Cell Weight*

The *C. roseus* CMCs were separated from the media by vacuum filtration and transferred to a Petri dish lined with a piece of filter paper. The cells were dried to constant weight at 50 ◦C and weighed on a ten-thousandth scale. All samples were measured in triplicate.

#### *4.5. Alkaloid Extraction and Determination*

#### 4.5.1. Chemicals

Vindoline, catharanthine, and ajmalicine were HPLC grade and were purchased from Aladdin (Aladdin Reagents Co., Shanghai, China), and all other chemicals were of analytical grade.

#### 4.5.2. HPLC-MS Analysis of Alkaloids

The alkaloid extraction from 6-day-old suspensions of *C. roseus* CMCs was performed according to the methods of Wang [31]. Alkaloid extracts were dissolved in 1.0 mL of methanol and analyzed by HPLC-MS/MS according to the methods of Ferreres et al. [32] with slight modifications.

The HPLC system was equipped with an Agilent 1260 series system (Agilent Technologies, Santa Clara, California, USA). Chromatographic separations were carried out on a 250 mm × 4.6 mm, 5 μm, Phenomenex Gemini C18 column (Phenomenex Inc., Torrance, CA, USA) at 25 ◦C. Elution was performed with a flow rate of 1 mL/min. Solvents used were acetonitrile (A) and acetic acid 1% (B). The gradient was as follows: 12% A at 0 min, 20% A at 30 min, 50% A at 40 min, 50% A at 45 min, 12% A at 47 min, 12% A at 55 min and the injection volume 10 μL. Spectroscopic data from all peaks were accumulated in the range 240–400 nm, and chromatograms were recorded at 255, 280, 290, 306, and 320 nm.

The mass detector was an X500R QTOF mass spectrometer (AB Sciex Pte. Ltd., Redwood City, CA, USA) equipped with an electrospray ionization (ESI) system and controlled by SCIEX OS 1.3.1 software (AB Sciex Pte. Ltd., Redwood City, CA, USA). The experimental parameters were as follows: curtain gas, 30 psi; ion source gas, GS1 60 psi and GS2 60 psi; temperature, 550 ◦C; ion spray voltage, 5.5 kV; collision gas, 7 psi. Other parameters were as follows: declustering potential (DP), 80 V; collision energy (collision energy spread), 10 V (0 V), and 35 V (15 V). The mass spectrometer was operated in the full-scan mode in the *m*/*z* range of 100–1500. The information-dependent acquisition spectra were automatically performed with helium as the collision gas in the *m*/*z* range of 50–1500. MS data were acquired in positive ionization mode. The isolation width of the parent ions for the following MS fragmentation events was set at ±50 mDa.

## 4.5.3. Quantitative Analysis of Alkaloids

Alkaloid extraction was performed according to the methods of Wang et al. [31]. Alkaloid extracts were dissolved in 1.0 mL of methanol and analyzed by HPLC and ion-pair extraction-spectrophotometry as described by Yang et al. [33].

The contents of vindoline, catharanthine, and ajmalicine were determined at 25 ◦C by HPLC analysis using an Agilent 1260 series system (Agilent Technologies, Santa Clara, CA, USA) and a Phenomenex Gemini C18 column (250 mm × 4.6 mm, 5 μm) (Phenomenex Inc., Torrance, CA, USA). The mobile phase consisted of methanol/acetonitrile/10 mM ammonium acetate (15:40:45, *v*/*v*/*v*) at a flow rate of 1.0 mL/min. The detection wavelength was 280 nm, and the injection volume was 10 μL. Before injection, all samples were filtered with 0.45-μm nylon membrane filters (Jinteng Corp., Tianjin, China). Alkaloids were identified and quantified by comparing retention time and UV absorbance spectra with the standards. In a citric acid-phosphate buffer of pH 3.0, an ion-pair complex was formed between all alkaloids in the sample and the color reagent bromophenol blue upon 5 min of reaction at 30 ◦C. The complex was extracted with CHCl3 in which the absorbance of total alkaloids was measured at the wavelength of 413 nm. Vindoline was used as a reference standard in the preparation of the calibration curve. Each sample was analyzed in triplicate.

#### *4.6. Transcriptome Determination and Analysis*

#### 4.6.1. Sample Preparation

After pre-culture of *C. roseus* CMC and DDC suspension cultures for 6 day, 25 mg/L *A. flavus* mycelium elicitor was added to the CMCs (the experimental group, EG), and the EG was further cultured for another 48 h. The CMC check (CK) group and DDC group (the callus group, callus) were treated with an equal amount of sterile water.

#### 4.6.2. RNA Extraction, Library Preparation, and Transcriptome Sequencing

The cultures were frozen in liquid nitrogen and ground into powder with a mortar and pestle, and RNA extraction was performed as described by Liu et al. [34]. RNA purity was verified using the NanoPhotometer® spectrophotometer (IMPLEN, Los Angeles, CA, USA). The purity and concentration of the purified RNA were examined with a NanoPhotometer® spectrophotometer (IMPLEN, Los Angeles, CA, USA) and a Qubit® RNA Assay Kit in a Qubit® 2.0 Fluorometer (Life Technologies, Los Angeles, CA, USA).

A total amount of 1.5 μg RNA per sample was used as input material for RNA sample preparation. Sequencing libraries were generated using the NEBNext® UltraTM RNA Library Prep Kit for Illumina® (NEB) following the manufacturer's recommendations, and index codes were added to attribute sequences in each sample. PCR products were purified (AMPure XP system, Beckman Coulter Inc., Brea, CA, USA), and library quality was assessed on the Agilent Bioanalyzer 2100 system (Genomics & Bioinformatics Core Facility, Fort Wayne, IN, USA).

The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumina, Brea, CA, USA) according to the manufacturer's instructions. After cluster generation, the library preparations were sequenced on an Illumina Hiseq platform, and paired-end reads were generated.

## 4.6.3. De Novo Transcriptome Assembly and Gene Functional Annotation

Raw data (raw reads) in the FASTQ format were first processed through in-house Perl scripts. In this step, clean data (clean reads) were obtained by removing reads containing adapter and/or poly-N and low-quality reads from raw data.

Transcriptome assembly was accomplished based on the clean data using the Trinity de novo transcriptome assembly software (Trinity Technologies & software Solutions Pvt Ltd., Bengaluru, Karnataka, India) with min\_kmer\_cov set to 2 by default and all other parameters set to default in absence of a public reference genome for *C. roseus* [35].

All de novo-assembled unigenes were annotated based on the following databases: NCBI non-redundant protein sequences (NR, http://www.ncbi.nlm.nih.gov); NCBI non-redundant nucleotide sequences (NT, http://www.ncbi.nlm.nih.gov); Swiss-Prot (a manually-annotated and reviewed protein sequence database, http://www.expasy.ch/sprot); Protein family (Pfam, http: //xfam.org/) [36]; Gene Ontology (GO, http://www.geneontology.org); euKaryotic Ortholog Groups (KOG, http://www.ncbi.nlm.nih.gov) [37], and the Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.genome.jp/12eg) using BLASTX alignment with an *E*-value = 1e-10 [38,39].

The transcription factor families were identified and annotated using the iTAK software (IAITAM, Canton, OH, USA) [40,41].

#### 4.6.4. Differential Expression Analysis of Unigenes

Gene expression levels were estimated by RSEM [42] for each sample. Clean data were mapped back onto the assembled transcriptome. The read count for each gene was obtained from the mapping results.

Prior to differential gene expression analysis, the read counts were adjusted for each sequenced library using the edgeR program package through one scaling/normalization factor. Differential expression analysis of two samples was performed using the DEGseq R package [43,44]. The *p*-value was adjusted using the *q*-value [45]. *q*-value < 0.005, and |log2 (fold change)| > 1 was set as the threshold for significantly differential expression.

GO enrichment analysis of the differentially-expressed genes (DEGs) was implemented by the GOseq R package-based Wallenius non-central hyper-geometric distribution [46], which can adjust for gene length bias in DEGs.

KEGG is a database resource for understanding high-level functions and utilities of biological systems, such as the cell, the organism, and the ecosystem, from molecular-level information, especially large-scale molecular datasets generated by genome sequencing and other high-throughput experimental technologies [47]. KOBAS software (Peking University, Beijing, China) [48] was used to test the statistical enrichment of differential expression genes in KEGG pathways.
