Transcriptome dataset of Caenorhabditis elegans responses to varied microbial pathogens

Abstract

Transcriptome analysis through next-generation sequencing (NGS) is an invaluable tool for investigating changes in gene expression across diverse organisms. The nematode Caenorhabditis elegans (C. elegans) serves as an excellent model organism for dissecting host responses to bacterial infections. Here, our dataset obtained from bulk RNA-sequencing (RNA-seq) can be used to provide in-depth characterization of the mRNA transcriptome profiles of wild-type N2 animals and null mutants of the cytoskeletal regulatory gene unc-53/Nav2 following exposure to distinct bacterial environments: their natural laboratory food source, Escherichia coli OP50, the human and nematode pathogen Pseudomonas aeruginosa PA14, and the emerging pathogen Elizabethkingia anophelis Ag1. As proof of the dataset quality, downstream differential gene expression analysis reveals significant shifts in gene expression patterns within N2 and unc-53 mutants under varying bacterial conditions that will be useful for our companion studies investigating these pathways. These data provide an effective methodological framework for future investigators to investigate the interplay between cytoskeletal proteins and the innate immune response. The raw FASTQ files generated from our transcriptome experiment is deposited in the publicly available NCBI Sequence Read Archive (SRA) under the BioProject accession number PRJNA1010192, for further exploration and validation by the C. elegans research community.

Specifications Table

Subject	Transcriptomics
Specific subject area	Transcriptome analysis of C. elegans worms fed different diets
Data format	Raw data: fastq sequencing files
Type of data	Table, Chart, Graph, Fig.
Data collection	RNA sequencing, Illumina HiSeq 2 × 150bp
Data source location	Institution: Eastern Mennonite University City/Town/Region: Harrisonburg, VA Country: USA
Data accessibility	Repository name: NCBI SRA Data identification number: PRJNA1010192 Direct URL to data: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1010192 [1]
Description of data collection	PolyA+ RNA was sequenced from C. elegans fed different diets

1. Value of the Data

• •These data will be valuable to members of the C. elegans research community interested in characterizing transcriptional changes between wild-type animals exposed to bacterial pathogens.
• •These data will also be valuable to members of the worm research community interested in characterizing transcriptional changes between unc-53 mutant animals exposed to bacterial pathogens.
• •These data can be used in subsequent comparative genomics studies aimed to understand the underlying innate immune responses of nematodes to bacterial infections.

2. Background

Disruptions in innate immunity increase vulnerability to invading pathogens, promote dysregulated inflammatory responses, and contribute to the development of autoimmune diseases and cancer [2,3]. The soil nematode C. elegans has been used to uncover host responses to a diversity of microbes and has aided in the identification of novel molecular targets controlling innate immunity [4]. These data describe C. elegans transcriptional responses to infection by the emerging pathogen Elizabethkingia anophelis (Ea) and the putative role of the unc-53 gene on transcription during infection with pathogenic Pseudomonas aeruginosa PA14. While Ea has gained attention as an important bacterium responsible for community and hospital outbreaks [5,6], little is known about its pathogenesis, and useful animal models exploring Ea pathogenesis are limited [7]. unc-53 encodes a Neuron Navigator (NAV) family protein with roles in cell migration and cellular processes through its interaction with signaling proteins and the cytoskeleton [8,9]. Evidence from our lab indicates that unc-53 mutants are highly susceptible to P. aeruginosa, and NAV homologs are needed for a robust immune response [10,11]. While studies exploring unc-53 and the NAVs in immunity exist, there are no public datasets detailing the global expression responses of mutant unc-53/Nav infected animals.

3. Data Description

These data provide a methodological framework useful in the characterization of the signaling pathways implicated in understudied infectious diseases and to the identification of new host innate immunity genes and pathways. These analyses were conducted using Illumina mRNA-seq followed by an open source bioinformatics pipeline to ensure sequence quality and repeatable transcriptome analysis (Fig. 1). These data demonstrate transcriptional changes between wild-type and unc-53 animals exposed to various bacterial pathogens and will be of use to the nematode research community for studying innate immune responses to bacterial infections.

Fig. 1.

Sequence quality control and workflow of data analysis. (A) Pipeline of tissue collection from C. elegans to bioinformatic tools used for downstream data analysis. (B) Sequence counts histogram for each sample illustrating unique (blue) and duplicate (black) reads. (C) Per sequence Phred scores showing the high quality of the sequenced reads with all average scores > 30. (D) Per base Phred scores showing the high quality of sequenced base pairs with almost all samples being > 30.

4. Experimental Design, Materials and Methods

4.1. C. elegans strain maintenance and treatment

C. elegans strains N2 (wild-type) and unc-53 (n152) were obtained from the C. elegans Genetics Centre (CGC) and were maintained by standard protocols using E. coli OP50 as a food source [12]. unc-53 (n152) is a true null allele resulting from a 319 bp deletion spanning alternative exon 18 to intron 19 and underwent six rounds of backcrossing before experimentation. To prepare worms for bacterial treatment, large populations of N2 and unc-53 (n152) animals were separately grown on standard NGM-OP50 plates and were synchronized by egg-preparation in bleach followed by L1-arrest overnight in M9 buffer. L1 animals were grown to the L4 stage at 20°C (40 h). L4 animals were washed off growth plates using M9 buffer and subjected to three additional washes in M9 buffer to remove residual bacteria. Approximately 1000 L4 animals per replicate (six replicates total per treatment) were dropped onto previously prepared slow kill assay (SKA) plates containing either E. coli OP50 (obtained from the CGC) or P. aeruginosa PA14 (gift of the Ausubel laboratory) for 6 h at 20°C [13]. Similarly, replicates of N2 worms were also subjected to E. anophelis Ag1 (BEI Resources) for 6 h at 20°C. Worms were harvested into 15 ml conical tubes post-infection using M9 buffer and three additional M9 buffer washes were completed to remove additional bacteria followed by removal of the M9 buffer. Worm pellets were resuspended in 500-700µl of Trizol to a final volume of 800µl and stored at -80°C.

Transcriptome Analysis of RNA

Frozen worms underwent six rounds of freeze-cracking by alternating them between a dry-ice-ethanol bath and 37°C water bath. RNA extraction was conducted using a Zymo Direct-Zol RNA Miniprep kit (Zymo Research) following the manufacturer's protocol with a final elution in 45µl of DNAse/RNAse-Free water and were stored at -80°C before subsequent analysis. Total RNA quality and quantity was assessed using UV spectrophotometer and Agilent TapeStation analysis. High quality total RNA samples were sent to Azenta Life Sciences (Chelmsford, Massachusetts) for mRNA transcriptome analysis. Illumina stranded TrueSeq cDNA libraries were constructed using poly dT enrichment for each sample following the manufacturer's protocol. The resulting average size of the cDNA libraries was approximately 300bp. Libraries for cDNA samples were sequenced using the Illumina HiSeq 4000 platform yielding 22-34.4 million 150 bp paired end (PE) sequence reads per sample (Table 1). Raw sequence data was returned and analyzed using a previously established open access bioinformatics package (Fig 1a) [14].

Table 1.

RNA-seq samples, read metrics, and public SRA accessions.

Sample name	Sequencer	Treatment group	# of reads (M)	Uniquely mapped reads (M)	Uniquely mapped reads%	SRA accession #
A1	Illumina Hiseq 2 × 150	A	26.6	23.0	86.6	SRR25780695
A2	Illumina Hiseq 2 × 150	A	26.6	23.8	89.3	SRR25780694
A3	Illumina Hiseq 2 × 150	A	26.1	23.1	88.5	SRR25780683
A4	Illumina Hiseq 2 × 150	A	22	19.6	89.1	SRR25780675
A5	Illumina Hiseq 2 × 150	A	29.7	26.9	90.6	SRR25780674
B1	Illumina Hiseq 2 × 150	B	27.4	23.7	86.3	SRR25780673
B2	Illumina Hiseq 2 × 150	B	34.4	31.0	90.2	SRR25780672
B3	Illumina Hiseq 2 × 150	B	27.2	24.4	89.9	SRR25780671
B4	Illumina Hiseq 2 × 150	B	23.9	21.5	89.9	SRR25780670
B5	Illumina Hiseq 2 × 150	B	28.4	25.5	89.9	SRR25780669
C1	Illumina Hiseq 2 × 150	C	29.9	26.7	89.4	SRR25780693
C2	Illumina Hiseq 2 × 150	C	26.4	23.8	90	SRR25780692
C3	Illumina Hiseq 2 × 150	C	27.9	24.9	89.3	SRR25780691
C4	Illumina Hiseq 2 × 150	C	28	25.4	90.7	SRR25780690
C5	Illumina Hiseq 2 × 150	C	30.6	27.6	90.2	SRR25780689
C6	Illumina Hiseq 2 × 150	C	31.5	28.7	91	SRR25780688
D1	Illumina Hiseq 2 × 150	D	30.2	27.0	89.4	SRR25780687
D2	Illumina Hiseq 2 × 150	D	26.5	23.9	90	SRR25780686
D4	Illumina Hiseq 2 × 150	D	30.3	27.2	90	SRR25780685
D5	Illumina Hiseq 2 × 150	D	29.2	26.5	90.9	SRR25780684
D6	Illumina Hiseq 2 × 150	D	23.2	20.3	87.3	SRR25780682
E1	Illumina Hiseq 2 × 150	E	27.7	24.7	89	SRR25780681
E2	Illumina Hiseq 2 × 150	E	28.8	25.9	89.7	SRR25780680
E3	Illumina Hiseq 2 × 150	E	25	22.3	89.1	SRR25780679
E4	Illumina Hiseq 2 × 150	E	26.6	23.8	89.5	SRR25780678
E5	Illumina Hiseq 2 × 150	E	27	24.3	90	SRR25780677
E6	Illumina Hiseq 2 × 150	E	28.4	25.4	89.2	SRR25780676

4.2. Bioinformatics analysis and sequence alignment QC

The overview or our RNA-seq experiment is described in Fig. 1a. QC of raw Illumina data in the 54 FASTQ files was assessed with FastQC and MultiQC software (see Code Availability 1-2). Fig. 1c-d demonstrates that all 54 FASTQ raw files possess average per sequence and per base Phred score ≥28 respectively, indicating sufficient quality. Filtered reads were aligned to the C. elegans WBcel235 reference genome using HISAT2 (see Code Availability 3). Aggregate plots for HISAT2 alignment were generated with MultiQC (see Code Availability 2). An average of 24.9 million reads per sample mapped to the WBcel235 ranging from 86.3% to 91% of total input reads (Table 1; Fig. 2a). Taken together, these data denote high quality sequencing and mapping of mRNA reads sufficient for subsequent analysis of differential gene expression.

Fig. 2.

Overview of sequence alignment and data dispersion. (A) Sequence alignment histogram showing the number of HISAT2 paired ends (PE) properly (blue) or improperly (yellow/red) aligned to the C. elegans WBcel235 reference genome. (B) Multi dimensional scaling (MDS) plot illustrating the dispersion of each of the individual samples according to their variance in the two dimensions. (C) Volcano plot displaying up-regulated and down-regulated genes in the E. coli OP50 + N2 treatment compared to the P. aeruginosa + N2 treatment. These differentially expressed genes (DEGs) are plotted with respect to the log₂ fold change and -log₁₀ P values.

Differential gene expression analysis

HISAT2 alignments were fed into featureCounts software to quantify reads in each sample aligning to exonic sequences annotated in the WBcel235 C. elegans reference genome (see Code Availability 4). Gene count tables generated by featureCounts were then fed into the edgeR statistical package to determine variant in gene expression between samples (see Code Availability 5). edgeR was also employed to calculate a multidimensional scaling (MDS) plot explaining the variance between sample groups in addition to the similarity of replicates for all 27 samples (Fig. 2b). To specifically highlight the utility of this dataset for studying transcriptome variance associated with exposure of worms to pathogenic bacteria, differential gene expression output tables from control worms exposed to non pathogenic E. coli and pathogenic P. aeruginosa were used to create a volcano plot representing significant differentially expressed genes (DEGs) between these two samples groups in terms of -log10(false discovery rate) versus log2(fold change). As a proof of principle, this analysis demonstrated marked increases in the transcription of immune response genes in nematodes exposed to P. aeruginosa (Fig. 2c).

4.3. Sequencing read and alignment quality

FastQC and MultiQC software analyses plotting mean Phred quality scores demonstrates that QC metrics are well within the acceptable range for RNA-seq data analysis (Fig. 1c-d). FASTQ files contained 22-34.4 million high quality raw reads per sample (Fig 1.b; Table 1). 86.3–91% of these high quality raw reads were mapped to the C. elegans WBcel235 genome assembly (Fig. 2a, Table 1).

4.4. Code availability

The following programs were used for QC, data analysis and data plots as described in the main text:

• 1.FastQC, version 0.11.5 accessible in the CyVerse Discovery Environment for QC analysis of FASTQ sequencing data: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
• 2.MultiQC, version 1.11 accessible in the CyVerse Discovery Environment to aggregate FastQC and HISAT2 alignment data: https://multiqc.info
• 3.HISAT2-index-align-2.1 accessible in the CyVerse Discovery Environment to index and align FASTQ reads to the C. elegans WBcel235 fasta genome: http://daehwankimlab.github.io/hisat2/
• 4.featureCounts, version 1.6.0 accessible in the CyVerse Discovery Environment to assign sequence reads to exonic features of the C. elegans WBcel235.109 genome: https://subread.sourceforge.net/featureCounts.html#:~:text=featureCounts%20is%20a%20highly%20efficient,and%20genomic%20DNA%2Dseq%20reads.
• 5.edgeR, version 2.0 accessible in the CyVerse Discovery Environment to quantify and visualize differentially expressed transcripts across all treatments: https://bioconductor.org/packages/release/bioc/html/edgeR.html

All scripts used for QC and data analysis in this study are available at: https://github.com/JonathanGM70/Microbial_Pathogens_on_C.elegans

4.5. Usage notes

All of the computational analyses used to describe our data set outlined in Fig. 1b are a collection of freely available tools. These analyses, however, are interchangeable with many other currently available tools for achieving similar experimental outcomes. Our data can be aligned to any available C. elegans reference genome or transcriptome, including the most recent WBcel235 reference assembly using a variety of freely available aligners. In this study we used the HISAT2 for traditional gene expression analysis. Alternatively, much faster alignment-free transcriptome pseudoaligners such as Kallisto can be applied to these data with the specific intent of expression quantification of previously characterized mRNA isoforms [16]. Alignment-free pipelines reduce the complexity and data imprint required for analysis but are not suitable for novel isoform analysis as demonstrated in our group's previous work [17]. Here our gene quantification and differential gene expression analysis were carried out using featureCounts and edgeR, however other openly available softwares such as StringTie may also be used for similar analysis [15]. Importantly, analyses demonstrated in Figs. 1,2 show the high quality of our sequencing, read alignment and sampling strategy indicating this data set will be compatible with any transcriptome analysis tools that are currently available as well as new tools that may become available down the road.

This dataset characterizes wild-type C. elegans fed their natural laboratory food source E. coli OP50 in addition to the well-studied pathogen P. aeruginosa PA14, and the emerging pathogen E. anophelis Ag1. A previous transcriptome study investigated the effect of switching C. elegans to a high carbohydrate diet, however, this is one of the first RNA-seq investigation into a diet of pathogenic bacteria [18]. An experimental design that includes three points of comparison between transcriptional responses to different bacteria will prove especially valuable for future research. Despite their simplicity, C. elegans have evolved diverse pathogen-specific protective responses to microbial pathogens [4]. Given the abundance of publicly accessible genomics datasets using C. elegans, the potential arises to integrate our data with what is already available to understand the pathogenesis of these bacteria. This holds particular significance for E. anophelis, as viable pathogenic models are limited. In addition to wild-type worms we also exposed null mutant unc-53 (n152) worms to E. coli OP50 and P. aeruginosa PA14. These datasets can be used alongside cell biology and genetics tools to understand the immunoprotective role of UNC-53, a cytoskeletal protein with homology to the NAV family proteins with known roles in the innate immune response [10,11].

Limitations

For our experiment, it should be noted that RNAs were extracted from whole worms. Therefore, our analysis is representative of heterogeneous mixtures of cell types within these animals. Additionally, cDNA libraries were prepared using a poly dT primer. This dictates that our data is representative of polyadenylated transcripts but does not include many non-coding or other non-polyadenylated transcripts. Also, poly dT priming often introduces a 3’ bias for transcript characterization, particularly for long messages. Lastly, though the quantity of our reads depth in this study is below the threshold for rare isoform identification, there are enough reads per sample in our study to characterize abundant alternative mRNA isoforms as previously demonstrated [19]. Taking these considerations into account, these data will be a useful resource for the C. elegans community and others, providing a robust and accurate analysis of the polyadenylated transcriptional networks in nematodes exposed to varied bacterial pathogens.

Ethics Statement

All animal experiments were conducted under the provisions of the National Institutes of Health guide for the care and use of laboratory animals. Research projects involving invertebrates do not require any IACUC oversight or review and were deemed “exempt”.

CRediT authorship contribution statement

Abigail M. Kaufman: Investigation, Resources, Software, Writing – original draft, Writing – review & editing. Jonathan G. Miller: Formal analysis, Software, Visualization, Data curation, Writing – original draft, Writing – review & editing. Emilio Fajardo: Investigation, Resources. Cheyenne Suamatai'a-Te'o: Investigation, Resources. Ray A. Enke: Conceptualization, Methodology, Writing – original draft, Supervision, Writing – review & editing. Kristopher L. Schmidt: Conceptualization, Methodology, Writing – original draft, Supervision, Project administration, Funding acquisition, Writing – review & editing.

Data Availability

• RNA-seq of C elegans whole worms (Original data) (NCBI SRA).