Time course gene expression data in colon of mice after exposure to food-grade E171

We investigated gene expression responses in BALB/c mice exposed by gavage to 5 mg/kg bw/day of E171 for 2, 7, 14 and 21 days. Food additive E171 (titanium dioxide) has been shown to induce oxidative stress and DNA damage in vitro as well as facilitating growth of colorectal tumours in vivo. Full genome expression changes of the colon of mice were investigated by using Agilent SurePrint G3 mouse Gene exp 60kv2 microarrays slides. The data presented in this DiB include all differentially expressed for each time point with EntrezGeneID, gene symbols, gene names and Log2FC as well as genes included in pathways after over-representation analysis in ConsensusPathDataBase. The functions of these genes in relation to the colon were described in our associated article (Proquin et al., 2017 in press) [1]. Raw and normalized gene expression data are available through NCBI GEO (GEO accession: GSE92563).


Data
Titanium dioxide, referred to as E171, is used as a colouring agent in various types of food like sweets, cookies, coffee creamers, and salad dressings [1,2]. To establish molecular responses that may relate to potential health effects in the colon, BALB/c mouse were intragastrically exposed to 5 mg/kg bw/day of E171 for 2, 7, 14, and 21 days. Microarray analyses of the colon of the mice showed the effects of E171 exposure on the whole transcriptome. The number of DEG was 417 after 2 days of exposure (Table 1), 971 after 7 days of exposure (Table 2), 1512 after 14 days (Table 3), and 229 after 21 days of exposure ( Table 4). The data shows that exposure to E171 affects the expression of genes which are involved in oxidative stress, immune response, DNA repair, development of cancer for instance colon cancer, and regulation of GPCR/olfactory and serotonin receptors (Tables 5, 6, 7, and 8). A relatively large proportion of the DEG could not be linked to known molecular pathways which indicate that E171 is affecting the biological response beyond currently known processes (Figure 3, in Ref. [3]).

Mouse model
Thirty-two BALB/c mice (16 males, 16 females) of 4-6 weeks old (Harlan Laboratories, Mexico) underwent exposure to E171 exposure after ethical approval from the Comité de Ética de la Facultad de Estudios Superiores Iztacala de la Universidad Nacional Autónoma de México under the number: FESI-ICY-I151. They were housed in polycarbonate cages and kept in a housing room (12 h light/dark cycles, 50-60% relative humidity, air filtered until 5 µm particles and exchanged 18 times/h, and 21°C ). After one week of acclimation, the mice were randomly divided in 2 groups: the control group (8 males, 8 females) was exposed by gavage to sonicated sterile water (30 min 60 Hz), and the exposure group (8 males, 8 females) was exposed to 5 mg/kg bw/day of sonicated sterile E171 (30 min 60 Hz). Four mice (2 males, 2 females) were sacrificed in a humid chamber with sevoflurane after 2, 7, 14 and 21 days. Colons were collected and put overnight at 4°C in a tube containing RNAlater® (Thermofischer, The Netherlands). Next day the remaining RNAlater® was discarded. The colons were stored at −80°C until RNA isolation. Transportation of the samples for 2 days in dry ice at −80°C has been done by a specialised shipping company to the Department of Toxicogenomics, Maastricht University, Maastricht, the Netherlands. The shipping company monitored the temperature of the samples throughout the shipping process with a thermometer to ensure stable freezing conditions and optimal sample quality. Immediately upon arrival the samples have been put at −80°C.

RNA isolation
Total RNA was isolated from the distal part of colon. Colons were first submerged in Qiazol (Qiagen, The Netherlands) and subsequently disrupted and homogenized using a Mini Bead Beater (BioSpec Products, The Netherlands) (48 beats/s for 30 s). miRNeasy Mini Kit was used for RNA isolation (Qiagen, The Netherlands) including a DNase treatment, according to the manufacturer's protocol [4]. RNA concentrations were measured using a Nanodrop® spectrophotometer (Thermofischer, The Netherlands) and the integrity of total RNA was determined on a Bioanalyzer (Agilent Technologies, The Netherlands). RNA Integrity Number (RIN) higher than 6 was compulsory for each sample to be used for microarray analysis. It was the case for all samples with an average of 8.8 7 0.7. Table 5 Group of pathways, pathways, genes related to this pathways and log2FC values after 2 days of exposure to E171 in BALB/c mice. Numbers in bold are upregulated genes. Log2FC¼ Log2 fold change obtained with LIMMA script with correction for its own time-matched control.

Microarray preparation and data pre-processing
Samples were labelled with Cyanine 3 (Cy 3), hybridized, and washed according to the One-Color Microarray-Based Gene Expression Analysis protocol version 6.6 (Agilent Technologies, The Netherlands) [5]. Hybridization was performed on Agilent SurePrint G3 mouse Gene exp 60kv2 microarrays. Samples were scanned using an Agilent DNA Microarray Scanner with Surescan High-resolution Technology (Agilent Technologies, The Netherlands). Raw data was extracted and checked by the quality control pipeline provided by Agilent (Feature extraction software (FES) version 10.7.3.1). Another quality check was performed with an in-house pipeline in R (https://github.com/BiGCAT-UM/ arrayQC_Module/) and data was normalized with local background correction, flagging of bad spots, controls and spots with too low intensity, log2 transformation and quantile normalization. Raw data with expression values and genes were selected for data analysis based on flags and missing values (GEO accession: GSE92563). Height groups were defined: control 2 days, 7 days, 14 days, 21 days for the controls and E171 2 days, 7 days, 14 days, 21 days for the exposed samples. Within each group, unique identifiers had to pass for the spot and have less than 40% of missing values. In all groups, unique identifiers had to have an average expression 44, missing values were imputed by k-nearest neighbours (k-NN; k-value 15), and repeated identifiers merged on the median. Differentially expressed genes were extracted with a Linear Mixed Model Analysis for Microarrays (LIMMA) [6] (version 1.0) [7] with the following criteria: a fold-change (FC) of 1.5 and a p-value of 0.05. The data of each control time point (control) was corrected from the time-matched exposed mice to E171. DEG extracted are shown in Tables 1, 2, 3, and 4.

Pathway analysis
Functional annotation of these treatment-specific DEG was performed with Consensus Pathway Database (CPDB) for an over-representation gene set analysis (ORA) [8,9]. All the available databases from CPDB were used (release MM9, 11 Oct. 2013) with settings in the "pathways as defined by pathway databases" with a minimum overlap of input list of 2 and a p-value cut-off of p o0.01. For each annotation set, the p-value is calculated using a Fisher's exact test. CPBD corrects for multiple hypothesis testing using the false discovery rate procedure within each type of annotation set [8,9]. Results of the ORA are available in Tables 5, 6, 7, and 8.