A time-resolved RNA-sequencing dataset of transcriptional responses in PC12 cells to NGF withdrawal and replenishment

Abstract

Nerve Growth Factor (NGF) is a pleotropic extracellular signalling peptide with neurotrophic, cell differentiation, and cell survival functions. Binding of NGF ligand to tyrosine kinase receptors stimulates intracellular cascades to impact gene transcription. Transcriptional responses coordinate neurodevelopment, and regulate the sensitivity and excitability of populations of adult neurons. NGF is synthesized during inflammation, inducing plasticity of sensory neurons and contributing to chronic pain. PC12 cells, derived from rat pheochromocytoma, are a classical model of NGF responses, that differentiate upon NGF treatment into neuron-like cells, with neurites and growth cones dependent on continued exposure to NGF. This dataset comprises a time series from NGF-differentiated PC12 cells subject to NGF withdrawal and subsequent replenishment. These data serve as a resource for the community to elucidate NGF-dependent time-resolved gene transcription in peripheral neuron-like cells. RNA was extracted, sequenced, and mapped to the rat genome. QC measures and analysis of time-dependent gene expression changes validated by analysis of known marker genes show that this sequencing data is robust and contains thousands of transcriptional events for future study. These data are available in the sequence read archive (SRA) and serve as a valuable resource for the study of NGF-dependent transcription.

Specifications Table

Subject	Biology
Specific subject area	Transcriptional changes dependent on neurotrophic withdrawal and replenishment.
Type of data	Fastq files Raw
Data collection	Cells were cultured for a time series of NGF withdrawal and replenishment. RNA was isolated from cells at designated time points using Qiagen RNeasy Mini spin columns and sequenced using an Illumina HiSeq 2000. Data quality was assessed using FastQC software, and bad quality sequence removed using Trimmomatic. Reads were aligned to the Rat genome using STAR aligner, and differential expression analyses performed in R Studio using the DESeq2 package.
Data source location	The University of New England, Biddeford, Maine
Data accessibility	Repository name: NCBI Sequence Read Archiv Data identification number: BioProject accession: PRJNA1269504 Direct URL to data: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1269504
Related research article	none

1. Value of the Data

• •Nerve growth factor (NGF) signaling impacts diverse disease states ranging from responses to tissue injury and neuroregeneration, to after myocardial infarction and autonomic dysreflexia following spinal cord injury.
• •These data, from a rat PC12 pheochromocytoma cell model, profile transcriptional changes in response to NGF loss and subsequent re-exposure.
• •The timecourse design of this experiment allows differential gene expression analyses of fluctuations in NGF-dependent transcriptional responses over hours to days.
• •These data can be used to generate testable hypotheses about the impact of transcriptional responses to NGF availability on neuronal properties, that can be confirmed/validated using primary culture and in vivo models.
• •This dataset provides distinct clusters of hundreds of coregulated genes. These include clusters of transcripts that are restored when NGF is replenished, and clusters of transcripts that remain up or down-regulated after NGF re-exposure over this timecourse.

2. Background

NGF plays a fundamental role in the development of the mammalian nervous system. After a developmental window, the majority of CNS neurons and 50% of PNS neurons switch to non-NGF dependency [1]. In the adult, sustained NGF concentrations are required for homeostasis, including maintenance of projections of CNS cholinergic neurons [2]. Upon tissue injury, NGF synthesis promotes neuronal plasticity, including sprouting [[3], [4], [5]] and sensitization [6], and is required for wound healing and reinnervation [7]. NGF responses are associated with a variety of diseases including sudden cardiac death [8], autonomic dysreflexia following spinal cord injury [9], autoimmune conditions [10] and neuropathies [11]. Therefore, further understanding of responses to NGF availability may lead to the elucidation of neurodevelopment, plasticity and disease mechanisms.

Existing publicly available transcriptomic data from in vitro models have served as valuable resources for characterizing responses to NGF, including various concentrations of continuous treatments (1, 50 and 100ng/ml NGF) or short transient pulsatile treatments [[12], [13], [14]], demonstrating the utility of cell culture approaches. However, these datasets do not measure gene expression changes over hours to days. This dataset therefore provides a valuable resource for the identification of NGF-dependent transcripts and their regulation patterns following NGF deprivation and replenishment over a more prolonged time series [15].

3. Data Description

3.1. Data generation

PC12 cells are a robust model of NGF response. Upon NGF stimulation, PC12 cells undergo differentiation and develop a neuron-like phenotype, both morphologically and physiologically [4]. Withdrawal of NGF from these cells leads to changes in transcriptional pathways that result in neurite retreat and apoptosis, while replenishing NGF can rescue these effects [13]. While previous studies have published gene expression profiles from PC12 cells with NGF stimulation, there is no available expression profiling data on the effects of NGF depletion and replenishment over multiple days in PC12 cells.

We employed a timecourse of NGF withdrawal and exposure in PC12 cells, and isolated RNA from culture at multiple timepoints (see Fig. 1a). Cells were plated on collagen and cultured with NGF for 6 days to allow them to differentiate, at which point NGF was depleted from culture for 48 hours, and subsequently replenished for 24 hours. Baseline RNA samples were harvested from the cells after the initial 6 days of differentiation. Further RNA samples were taken after 8 and 48 hours of NGF withdrawal, and 8 and 24 hours of NGF replenishment, allowing for the temporal examination of resulting transcriptional changes.

Fig. 1.

RNA sequencing and quality control from PC12 cells following NGF withdrawal and replenishment. (a) Timeline of PC12 treatment and collection points. Cells were plated and cultured with NGF for 6 days before the baseline collection point. Further collections were taken 8h, 48h, 56h and 72h after baseline, with NGF withdrawn and replenished as indicated. (b) Read confidence as measured by FastQC, collated by MultiQC, for all samples both pre- and post-trimming. (c) Number of input reads and (d) uniquely mapped reads as determined by STAR. Error bars represent mean ± SD. (e) PCA was performed to visualize the variation among samples and conditions. The axes represent principal components PC1 and PC2, which account for 43% and 27% of sample variance, respectively.

3.2. RNA sequencing quality assessment

FASTQ files were subjected to quality control using the FastQC software. Per-base Phred scores indicate high confidence reads, which were improved by trimming residual Illumina adapters from the reads using Trimmomatic (Fig. 1b). Surviving read count and unique mapping rate were verified using the STAR aligner (Fig. 1c, d). Sample information is shown in Table 1, and all raw fastq files can be found on the short read archive here: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1269504

Table 1.

Sample identification information. All data files are available on the SRA.

SRA Sample Name	File Names	Time Point	Organism	Strain	Reads/ Sample
SAMN48789665	1_S1_R1_001.fastq.gz	Baseline	Rattus norvegicus	PC-12 CRL-1721	24333142
	1_S1_R2_001.fastq.gz
SAMN48789666	2_S2_R1_001.fastq.gz	Baseline	Rattus norvegicus	PC-12 CRL-1721	25035802
	2_S2_R2_001.fastq.gz
SAMN48789667	3_S3_R1_001.fastq.gz	Baseline	Rattus norvegicus	PC-12 CRL-1721	33830273
	3_S3_R2_001.fastq.gz
SAMN48789668	4_S4_R1_001.fastq.gz	8h	Rattus norvegicus	PC-12 CRL-1721	24532573
	4_S4_R2_001.fastq.gz
SAMN487896669	5_S5_R1_001.fastq.gz	8h	Rattus norvegicus	PC-12 CRL-1721	25957853
	5_S5_R2_001.fastq.gz
SAMN48789670	6_S6_R1_001.fastq.gz	8h	Rattus norvegicus	PC-12 CRL-1721	26074677
	6_S6_R2_001.fastq.gz
SAMN48789671	7_S7_R1_001.fastq.gz	48h	Rattus norvegicus	PC-12 CRL-1721	32394438
	7_S7_R2_001.fastq.gz
SAMN48789672	8_S8_R1_001.fastq.gz	48h	Rattus norvegicus	PC-12 CRL-1721	67193941
	8_S8_R2_001.fastq.gz
SAMN48789673	9_S9_R1_001.fastq.gz	48h	Rattus norvegicus	PC-12 CRL-1721	47149061
	9_S9_R2_001.fastq.gz
SAMN48789674	10_S10_R1_001.fastq.gz	56h	Rattus norvegicus	PC-12 CRL-1721	23600985
	10_S10_R2_001.fastq.gz
SAMN48789675	11_S11_R1_001.fastq.gz	56h	Rattus norvegicus	PC-12 CRL-1721	34586863
	11_S11_R2_001.fastq.gz
SAMN48789676	12_S12_R1_001.fastq.gz	56h	Rattus norvegicus	PC-12 CRL-1721	23528613
	12_S12_R2_001.fastq.gz
SAMN48789677	13_S13_R1_001.fastq.gz	72h	Rattus norvegicus	PC-12 CRL-1721	41246905
	13_S13_R2_001.fastq.gz
SAMN48789678	14_S14_R1_001.fastq.gz	72h	Rattus norvegicus	PC-12 CRL-1721	7960195
	14_S14_R2_001.fastq.gz
SAMN48789679	15_S15_R1_001.fastq.gz 15_S15_R2_001.fastq.gz	72h	Rattus norvegicus	PC-12 CRL-1721	22272018

3.3. Verification of transcriptional responses to NGF

3.3.1. Principal component analysis (PCA)

PCA was performed to reduce the dimensionality of the data and visualize variation between samples. The clustering of samples indicate consistency between treatment groups and a robust effect of NGF on gene expression (Fig. 1e).

3.3.2. Coincidence with established marker genes

Gene counts from the STAR aligner were imported into RStudio for differential analysis with DESeq2. Withdrawal timepoints were analyzed in reference to baseline expression, while replenishment timepoints were in reference to the 48h withdrawal timepoint.

Normalized gene counts were used to verify NGF responses in these cells. Syn1 [16], Tubb3 [17], Egr1 [18] and Map2 [19] have all been identified as marker genes dependent on NGF signaling in PC12 cells. In concordance, we observed these genes are downregulated and re-upregulated in response to NGF depletion and replenishment, validating NGF responses over this time series (Fig. 2a).

Fig. 2.

Validation of transcriptional response (a) Expression profiles of select NGF responsive marker genes responsive to NGF during NGF withdrawal. Error bars represent the mean ± SD; p-values are indicated on significance brackets, as determined by Tukey's post hoc test following ANOVA. Differential expression profiling revealed thousands of significant transcriptional changes. (b) UpSet graph visualizing the intersection of differentially expressed genes (DEGs) between timepoints. Bars to the left represent the total number of statistically differentially expressed genes (q<0.05) at each time point. The bars on top represent the number of DEGs represented in the group intersections indicated below. (c) 8 most significantly enriched expression profiles resulting from STEM clustering analysis. The number of genes associated with each expression profile is indicated above the models. The black line indicates the expression profile represented and red lines indicate individual gene expression patterns from the data. The p-value in the bottom left of each box indicates the statistical significance of the overrepresentation of each expression profile.

3.3.3. Verification of transcriptional changes over time

To verify that the data captured a robust transcriptional response to NGF, genes with statistically significant differential expression at any timepoint were compared using an UpSet [20] graph (Fig. 2b). First, significantly differentially expressed genes (DEGs) at each timepoint were split into 2 groups, upregulated and downregulated. This graph indicates the total number of DEGs in each of these groups, as well as the number of DEGs that are shared between groups, visualizing the potential use of this dataset for identification of NGF targets.

To determine if responses to NGF withdrawal are re-established, normalized read counts for all genes were exported to the Short Time-series Expression Miner (STEM) for analysis of expression profiles over the timecourse. Statistically overrepresented expression profiles are indicated in Fig. 2c, with the number of genes exhibiting each expression profile shown above. These expression profiles are generally associated with the timecourse of NGF exposure and represent the utility of this data in isolating genes that are correlated, positively or negatively, with NGF signaling in PC12 cells. In addition, this analysis demonstrates the potential of this data to identify transcripts that may remain elevated (e.g., transcripts in profiles 560, 598, 403) or decreased (e.g., profiles 545, 528) following replenishment.

4. Experimental Design, Materials and Methods

PC12 cells were purchased from ATCC (CRL-1721) and maintained per manufacturer protocol in growth media containing RPMI01640 media (ATCC, 30-2001) supplemented 10% Heat inactivated horse serum (ATCC, 30-2004) and 5% fetal bovine serum and supplemented with 1% penicillin/ streptomycin (Lonza BioWhittaker, 17-602E) and Gentamycin (VWR, 0304-10G).

Cells were plated on collagen-coated 12-well plates (Corning, 354400) in differentiation media (RPMI-1640, 1% Heat inactivated horse serum, Gibco, 26050070) and supplemented with 50mg/ml NGF (Sigma, N-6009). Fresh differentiation media with NGF was replenished every 2 days, for total of 6 days before the start of collections. For the 8 and 48 hour timepoints, cells were washed two times and supplemented with NGF free, low-serum media for 8 and 48 hours respectively before they were collected for RNA sequencing. Two more groups were replenished with NGF supplemented differentiation media for 8 and 24 hours following withdrawal (Fig. 1a). RNA was isolated from samples at each collection point using RNeasy Mini spin columns (Qiagen, 74104) and stored at -80°C until shipment. Three biological replicates were generated for each timepoint.

Rationale for timing of NGF treatments: Previous studies have explored the impact of immediate early responses to NGF signalling and withdrawal on cell differentiation [13]. This study attempts to explore NGF responses at prolonged timepoints, including the effects of re-exposure. Therefore, cells were first differentiated with NGF for 6 days to establish a stable neuronal transcriptional state. This was proceeded by an 8 hour and 48 hour NGF withdrawal period to model loss of trophic support over hours (“acute”) to days (“chronic”).

Sequencing was performed at the University of Delaware DNA Sequencing & Genotyping Center. Supplied RNA was sequenced via polyA selection using Illumina HiSeq 2000, PE 2×150. Adapters were trimmed from reads at the vendor prior to our receival of the data. Read quality was analyzed using FastQC v0.11.9 and MultiQC v1.9 (Fig. 1b), and we decided to perform a second trim. TruSeq3-PE adapters were trimmed from sequence reads using Trimmomatic v0.38. Reads were mapped to the mRatBN7.2 reference genome using the STAR aligner v2.7.11, and raw reads determined using the GeneCounts output. Differential gene expression analysis was performed using the DESeq2 package v1.46.0 on R v4.4.0. Principal component analysis was also performed using DESeq2, after regularized log normalization of the raw count data (Fig. 1e). The R code used to analyze the gene count data is publicly available at: https://github.com/pkneufeld/Neufeld2025_DESeq2/blob/main/PC12timecourse_DESeq2_upload.R

Normalized gene counts for all genes were exported from DESeq2 and loaded into the Short Time-series Expression Miner (STEM) v1.3.13. Reads were normalized to the baseline expression and a total of 20 expression model profiles were generated. Only significantly enriched expression profiles were selected for representation (Fig. 2c).

Limitations

Time course limitations: We sampled time points that likely account for a large proportion of but not all gene expression changes following NGF depletion and replenishment in PC12 cells. Future studies with increased resolution of sample points and/or a prolonged time series may therefore lead to further discovery of additional genes/profiles. Note however that NGF depletion causes cell stress and death in serum depleted conditions, limiting the potential withdrawal time. Also, our validation assessments demonstrated that this data captures thousands of gene expression changes, including coincidence with known NGF-response genes (Fig. 2a). These differential expression events could therefore be informative for future investigation at additional timepoints.

Reference to baseline: Differential expression analyses of withdrawal timepoints are referenced to 6 days in NGF (baseline), where cells are stably differentiated. This dataset does not therefore account for gene expression responses to longer times in culture with NGF. However, transcriptional responses to prolonged time in culture with NGF past 6 days are likely minimal compared to the robust differential expression signals seen upon NGF withdrawal and re-exposure(Fig. 2).

Cell line: PC12 cells are an in vitro model, from Rat pheochromocytoma, for cost effective and efficient generation of testable hypotheses about the impact of transcriptional responses to NGF. Therefore, responses may diverge from cultured primary neurons. In addition, this model does not capture tissue interactions or the impact of inflammatory mediators on disease processes. Also, non-compartmentalised cultures cannot differentiate between impacts of NGF on at nerve endings compared to cell bodies. The utility of this dataset lies in the identification of clusters of co-regulated transcripts and their expression patterns, but all targets should be validated in physiologically relevant models.

Ethics Statement

The authors have read and follow the ethical requirements for publication in Data in Brief and confirm that the current work does not involve human subjects, animal experiments, or any data collected from social media platforms.

CRediT Author Statement

All authors were involved in Conceptualization, Methodology, Data Curation, Writing, Visualization, Investigation, Reviewing and Editing.

Data Availability

• NCBI SRATranscriptional changes dependent on neurotrophic withdrawal and replenishment. (Original data).