Dataset of transcriptomic changes that occur in human preadipocytes over a 3-day course of exposure to 3,3',4,4',5-Pentachlorobiphenyl (PCB126).

Exposure to polychlorinated biphenyls (PCBs) has been associated with the development of metabolic syndrome, a cluster of diseases that includes obesity, diabetes, liver steatosis, and cardiovascular problems. PCBs accumulate and fat and are known to act on adipocytes and their precursors, termed preadipocytes. The PCB congener, PCB126, has been shown to activate the aryl hydrocarbon receptor (AhR) as well as proinflammatory genes. Here, we used RNAseq to assess gene transcript changes that occur in PCB126-exposed human preadipocytes over a time course. RNA was collected from 4 replicates of PCB126-exposed and control-treated preadipocytes at 9 h, 24 h, and 72 h post-exposure. RNA was processed for RNAseq analysis using a NovaSeq 6000 with an obtained minimum of 25 million paired-end 50 bp reads per sample. Reads were aligned using the salmon aligner and transcript expression values were summarized to the gene level using tximport. Gene transcript level counts comparing treated- versus control-treated cells were used for differential expression analysis using DESeq2. Differential expression Excel tables (one for each time point) were generated displaying average differential expression (log2 fold change) of the 4 replicates of treated versus control samples with cutoffs of 0.3 log2 fold change (increase or decrease) and p-values of less than 0.05. FastQ, raw, and differential expression tables were uploaded to GEO. A heat map of genes that were changed in common across all time points was generated using GraphPrism. The data generated from this analysis provides a full transcriptional profile of changes that occur over time in preadipocytes that have been exposed to PCB126. The rich datasets can be mined by other researchers to understand how PCB126 and other dioxin-like compounds, including other PCB congeners such as PCB77 and PCB118, affect biological pathways in preadipocytes and other cell types to cause disease.

Specifications Table

Value of the Data

The data provided here represents the first RNAseq data reported for exposure of human cells to PCB126, an important dioxin-like persistent organic pollutant known to activate the aryl hydrocarbon receptor (AhR).

The data provides a full transcriptional profile of changes that occur over time in preadipocytes that have been exposed to PCB126. Changes that occur over time could reveal novel pathways that are activated directly or secondarily upon exposure to PCB126.

Toxicologists and molecular biologists will be able to mine this data to unravel the cascade of gene pathways that are activated and altered upon acute exposure to PCB126 to help identify how and why PCB126 contributes to metabolic disease progression.

Datasets could be used to generate hypothesis for how other dioxin-like compounds such as PCB118 and PCB77 alter preadipocyte gene expression.

Gene expression changes found here in human preadipocytes could be compared to other experiments that used primary or immortalized rodent cell lines to study PCB126. These studies could provide important insight into how different species respond similarly or differently to PCB126 exposure.

The gene expression changes found to be elicited by PCB126 exposure of human preadipocytes provide clues as to how exposure to PCB126 or other dioxin-like compounds disrupts adipose function to cause metabolic syndrome.

Data Description

Normal human preadipocytes (NPAD) derived from subcutaneous adipose tissue were exposed to 10 µM PCB126 or DMSO over a time course and then subjected to RNA-Sequencing. The 10 μM concentration was chosen based on our previous studies demonstrating that this concentration was non-cytotoxic but caused significant inhibition of adipogenesis [2]. The specific time points of 9 h, 24 h, and 72 h were chosen to define the temporal changes in gene expression that occur after PCB126 exposure. This time frame is based on our previous findings demonstrating induction of the cytochrome p450 genes at the early time point of 9 h with a delayed inflammatory response that started at 24 h that was near maximal by 72 h [3]. Raw data obtained from an Illumina NovaSeq 6000 sequencer was converted to fastq format before being deposited in the Gene Expression Omnibus (GEO) database (Accession number: GSE193578). Table 1 outlines each raw data file and describes the treatment condition, either DMSO or 10 µM PCB126, as well as the duration of exposure to the treatment condition, 9-72 h. In addition to the raw counts in Table 1, the data was aligned to hg38 reference genome and the number of aligned reads and number of genes identified for each sample is reported in Table 2. After alignment, differentially expressed genes (DEG) were identified by comparing the 10 µM PCB126 treated NPAD data to the DMSO treated NPAD data for each of the 3 exposure durations. The list of DEG was filtered to only include genes that showed a log fold change ≥ |0.3| & adjusted p-value ≤ 0.05. Lists of raw counts for every gene and filtered DEG for each exposure duration and their corresponding log fold change and p-value are available in the files listed in Table 3 and can be found on GEO Accession number: GSE193578.

Table 1

List of accession number for each transcriptome in GEO database.

Sample	Treatment Condition	Exposure Duration	GEO Accession Number
Veh_1_9H	DMSO	9 h	GSM5814526
Veh_2_9H	DMSO	9 h	GSM5814527
Veh_3_9H	DMSO	9 h	GSM5814528
Veh_4_9H	DMSO	9 h	GSM5814529
X126_1_9H	10 µM PCB126	9 h	GSM5814530
X126_2_9H	10 µM PCB126	9 h	GSM5814531
X126_3_9H	10 µM PCB126	9 h	GSM5814532
X126_4_9H	10 µM PCB126	9 h	GSM5814533
Veh_1_Day1	DMSO	24 h	GSM5814534
Veh_2_ Day1	DMSO	24 h	GSM5814535
Veh_3_ Day1	DMSO	24 h	GSM5814536
Veh_4_ Day1	DMSO	24 h	GSM5814537
X126_1_ Day1	10 µM PCB126	24 h	GSM5814538
X126_2_ Day1	10 µM PCB126	24 h	GSM5814539
X126_3_ Day1	10 µM PCB126	24 h	GSM5814540
X126_4_ Day1	10 µM PCB126	24 h	GSM5814541
Veh_1_Day3	DMSO	72 h	GSM5814542
Veh_2_ Day3	DMSO	72 h	GSM5814543
Veh_3_ Day3	DMSO	72 h	GSM5814544
Veh_4_ Day3	DMSO	72 h	GSM5814545
X126_1_ Day3	10 µM PCB126	72 h	GSM5814546
X126_2_ Day3	10 µM PCB126	72 h	GSM5814547
X126_3_ Day3	10 µM PCB126	72 h	GSM5814548
X126_4_ Day3	10 µM PCB126	72 h	GSM5814549

Table 2

Summary statistics of reads mapping for each sample after alignment.

Sample	# of Mapped Reads	# of Genes Mapped
Veh_1_9h	29,352,009	20,093
Veh_2_9h	26,035,064	19,695
Veh_3_9h	40,601,343	20,699
Veh_4_9h	29,735,917	19,934
X126_1_9h	33,029,096	20,172
X126_2_9h	31,377,156	20,089
X126_3_9h	33,340,380	20,066
X126_4_9h	32,423,386	20,179
Veh_1_Day1	29,733,576	20,070
Veh_2_Day1	36,013,685	20,586
Veh_3_Day1	33,340,688	20,388
Veh_4_Day1	35,876,505	20,442
X126_1_Day1	35,647,938	20,429
X126_2_Day1	38,586,351	20,386
X126_3_Day1	33,551,975	20,100
X126_4_Day1	34,912,103	20,194
Veh_1_Day3	29,804,511	20,198
Veh_2_Day3	31,061,250	20,422
Veh_3_Day3	34,618,101	20,645
Veh_4_Day3	25,673,758	19,992
X126_1_Day3	33,644,941	20,143
X126_2_Day3	32,968,110	20,236
X126_3_Day3	32,465,246	19,879
X126_4_Day3	29,879,815	19,840

Table 3

Processed data files after alignment and differentially expressed gene analysis.

File Name	Description of Analysis	Exposure Duration
GSE193578_processed_raw_counts_all_genes_ vehicle_PCB126.csv.gz	Raw Counts after alignment	All
GSE193578_IPG-DEG_pcb126_9hour_vs_veh_9hour.xlsx	Differential Gene Expression between DMSO and PCB126 treated cells. Filtered to include genes with log fold change ≥ |0.3| & p-value ≤ 0.05	9 h
GSE193578_IPG-DEG_pcb126_dayone_vs_veh_dayone.xlsx		24 h
GSE193578_IPG-DEG_pcb126_daythree_vs_veh_daythree.xlsx		72 h

DEGs that had a p<0.05 and ≥0.3 log2 fold change comparing PCB126 treated to DMSO treated at all time points were selected, and a heatmap of their log fold change was generated showing those same genes for all 3 exposure durations (Fig. 1). Seventy-three genes were significantly upregulated and eighteen significantly downregulated at all three time points. The majority of these genes exhibited progressive upregulation or downregulation, respectively, over time. Many of the genes that were progressively upregulated are AhR-responsive genes (e.g., CYP1B1, TIPARP, CYP1A1) or proinflammatory genes (e.g., IL1B, IL1A, LIF).

Fig. 1

Heatmap of DEG changes over time in PCB126 treated preadipocytes compared to DMSO treated preadipocytes. DEG with a p < 0.05 at all exposure durations were selected, and their log fold change was plotted in a heat map to show the progression of gene changes from 9-72 h. Color map represents log fold change, with a negative fold change representing a decrease in gene expression and a positive fold change representing an increase in gene expression in PCB126 exposed cells compared to DMSO exposed cells.

Experimental Design, Materials and Methods

Human Subcutaneous Preadipocyte Exposure and RNA-Extraction

For this dataset, an immortalized human preadipocyte line (NPAD) was used for all conditions. The cells were originally isolated from subcutaneous fat harvested from a non-diabetic female donor. The isolated cells were then immortalized and characterized as previously described [4]. NPADs were cultured on tissue culture plastic in Preadipocyte Growth Media (PGM2, Lonza) throughout the duration of the treatment. For treatment, 80,000 cells were plated in a 6-well plate and cultured until 90% confluent. Once near confluent, the media was changed to PGM2 supplemented either with dimethyl sulfoxide (DMSO) or 10 µM PCB126 (C12H5Cl5; InChI: 1S/C12H5Cl5/c13-8-2-1-6(3-9(8)14)7-4-10(15)12(17)11(16)5-7/h1-5H; InChI Key: REHONNLQRWTIFF-UHFFFAOYSA-N; Canonical SMILES: 1 = CC(=C(C=C1C2 = CC(=C(C(=C2)Cl)Cl)Cl)Cl)Cl) dissolved in DMSO. The level of DMSO was held constant in all conditions at a level of 0.1% (v/v). The media remained on the cells until RNA harvesting. At each harvesting time point, the media was removed and cells were harvested using 1 mL of TRIzol Reagent (Invitrogen). Cells and TRIzol were pipetted several times to homogenize the sample and then transferred to Eppendorf tubes for further processing using a Qiagen RNeasy kit to isolate RNA for each sample. Each treatment condition and treatment duration was repeated 4 times to provide biological replicates.

RNA Library Preparation and Sequencing

To prepare RNA libraries for sequencing, 500 ng of RNA from each sample were incubated with oligo(dT) primer coated beads to enrich for polyA-containing transcripts. The enriched RNA pool was then fragmented, converted to cDNA, and ligated to indexes using an Illumina TruSeq stranded mRNA preparation kit (Cat. #RS-122-2101, Illumina). An Agilent Bioanalyzer 2100 was used to measure the molar concentrations of the indexed libraries and samples were pooled for sequencing. The concentration of each pool was measured using an Illumina Library Quantification kit (KAPA Biosystems) and sequenced on an Illumina NovaSeq 6000 genome sequencer at the Iowa Institute of Human Genetics. A minimum of 25 million paired-end 50 bp reads per sample were obtained.

Data Processing and Differential Expression Analysis

Reads were then converted from the native Illumina BCL format to fastq and processed using the ‘bcbio-nextgen’ pipeline (https://github.com/chapmanb/bcbio-nextgen). For alignment, the hg38 human genome was used as a reference and reads were aligned using hisat2 aligner [5,6]. For all samples ∼95% of RNA-seq reads uniquely mapped to the reference genome with ∼90% of mapped reads residing in an exonic region. For quality control, the ‘bcbio-nextgen’ pipeline runs MultiQC and returned no significant quality control problems for any of the samples reported here. In addition, the salmon aligner was used to quantify reads to the human transcriptome (GENCODE 39). Salmon transcript expression values were mapped to the gene level count estimates using tximport as described in (http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#transcript-abundance-files-and-tximport-tximeta) [7]. Non-normalized gene-level counts were then used as input to the DESeq2 model to identify differentially expressed genes (DEG) between DMSO and PCB126 treated NPADS at each exposure duration (DESeq2 performs within- and between-sample normalization internally) [8]. The code to reproduce the analysis can be found here: https://github.com/mchimenti/klingelhutz_rnaseq_july2020_pcb126. Excel tables displaying differentially expressed genes (log fold change ≥ |0.3| & adjusted p-value ≤ 0.05) comparing PCB126-treated versus DMSO-treated samples (averaged across 4 replicates of each) at the same time point were generated. The adjusted p-value was obtained using the Benjamini-Hochberg Method as implemented in DESeq2. A heatmap showing common differentially expressed genes across all time points was generated using GraphPrism.

Ethics Statement

The work described in this manuscript adheres to ethical publishing standards. The work did not involve human subjects. The immortal NPAD cell line used in this study has been published [4] and was derived from de-identified primary preadipocytes purchased from Lonza and obtained by consent.

CRediT authorship contribution statement

Francoise A. Gourronc: Conceptualization, Methodology, Investigation, Data curation, Writing – original draft. Brynn K. Helm: Investigation. Larry W. Robertson: Resources, Funding acquisition. Michael S. Chimenti: Data curation, Writing – original draft. Hans-Joachim Lehmler: Resources, Funding acquisition. James A. Ankrum: Conceptualization, Writing – original draft, Funding acquisition. Aloysius J. Klingelhutz: Conceptualization, Methodology, Investigation, Data curation, Writing – original draft, Supervision, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Subject	Health, Toxicology and Mutagenesis
Specific subject area	Temporal gene expression changes in preadipocytes caused by exposure to PCB126
Type of data	TableFigure
How the data were acquired	Data was acquired by performing RNA-sequencing using an Illumina NovaSeq 6000 genome sequencer.
Data format	Raw -FastqAnalyzed – “Raw Counts after alignment”Filtered Differential Gene Expression
Description of data collection	Immortalized normal human preadipocytes (NPAD) from a non-diabetic female donor were plated and cultured until 90% confluent. The cells were then treated with either 10 µM PCB126 or DMSO as a vehicle control. After 9, 24, and 72 h, cells were harvested for RNA-sequencing analysis. 4-replicates of each condition were collected.
Data source location	• Institution: University of Iowa • City/Town/Region: Iowa City, Iowa • Country: USA
Data accessibility	Repository name: Gene Expression Omnibus (GEO)Data identification number: GSE193578Direct URL to data: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE193578Code to reproduce the DEG analysis: https://github.com/mchimenti/klingelhutz_rnaseq_july2020_pcb126
Related research article	F.A. Gourronc, B.K. Helm, L.W. Robertson, M.S. Chimenti, H.J. Lehmler, J.A. Ankrum, A.J. Klingelhutz, Transcriptome sequencing of 3,3′,4,4′,5-Pentachlorobiphenyl (PCB126)-treated human preadipocytes demonstrates progressive changes in pathways associated with inflammation and diabetes., Toxicology in Vitro, Volume 83, 2022, 105396,https://doi.org/10.1016/j.tiv.2022.105396.[1]