Mitogen-induced transcriptional programming in human fibroblasts

Kiran L. Sharma a, 1, Shuo Jia a, 1, Tasnim H. Beacon a, b, Ifeoluwa Adewumi a, Camila Lo´pez a, Pingzhao Hu a, b, Wayne Xu a, b, James R. Davie a, b,*
a Department of Biochemistry and Medical Genetics, University of Manitoba, Winnipeg, Manitoba R3E 0J9, Canada
b CancerCare Manitoba Research Institute, CancerCare Manitoba, Winnipeg, Manitoba R3E 0V9, Canada
A R T I C L E I N F O

Keywords:
Immediate-early genes
Mitogen- and stress-activated protein kinase Protein kinase A
Transcriptome RNA sequencing

A B S T R A C T

Treatment of serum-starved quiescent human cells with fetal bovine serum (FBS), epidermal growth factor (EGF), or the phorbol ester (12-O-tetradecanoylphorbol-13-acetate, TPA) activates the RAS-MAPK pathway which initiates a transcriptional program which drives cells toward proliferation. Stimulation of the RAS-MAPK pathway activates mitogen- and stress-activated kinases (MSK) 1 and 2, which phosphorylate histone H3 at S10 (H3S10ph) or S28 (H3S28ph) (nucleosomal response) located at the regulatory regions of immediate-early genes, setting in motion a series of chromatin remodeling events that result in transcription initiation. To investigate immediate-early genes regulated by the MSK, we have completed transcriptome analyses (RNA sequencing) of human normal fibroblast cells (CCD-1070Sk) stimulated with EGF or TPA ± H89, a potent MSK/
PKA inhibitor. The induction of many immediate-early genes was independent of MSK activity. However, the induction of immediate-early genes attenuated with H89 also had reduced induction with the PKA inhibitor, Rp- cAMPS. Several EGF-induced genes, coding for transcriptional repressors, were further upregulated with H89 but not with Rp-cAMPS, suggesting a role for MSK in modulating the induction level of these genes.

Abbreviations:
EGF, epidermal growth factor; TPA, 12-O-tetradecanoylphorbol-13-acetate; MSK, mitogen- and stress-activated kinase; MAPK, mitogen-activated protein kinase; ERK, extracellular-signal-regulated kinase; H3S10ph, histone H3 phosphorylated at Ser 10; H3S28ph, histone H3 phosphorylated at Ser 28; RNASeq, RNA sequencing.

* Corresponding author at: Department of Biochemistry and Medical Genetics, University of Manitoba, 745 Bannatyne Avenue, Room 333A, Winnipeg, Manitoba, Canada.
E-mail address: [email protected] (J.R. Davie).
1 Both authors contributed equally to this work.
https://doi.org/10.1016/j.gene.2021.145842 Received 1 June 2021; Accepted 13 July 2021
Available online 16 July 2021
0378-1119/© 2021 Elsevier B.V. All rights reserved.

1. Introduction

Stimulation of serum-starved quiescent human fibroblasts with growth factors results in waves of gene transcription with the earliest expressed genes being the immediate-early genes, many of which encode transcription factors (FOS, JUN) (Fowler et al., 2011; Iyer et al., 1999). The expression of immediate-early genes is independent of pro- tein synthesis and occurs within an hour following stimulation (Iyer et al., 1999). The analysis of the transcriptional response of human foreskin fibroblasts to serum led to the discovery that many of the serum-induced genes were involved in wound healing (Iyer et al., 1999). Subsequently this “wound-response signature” and a subset of these induced genes (the “core serum response”) were found to predict cancer progression and cancer survival in breast cancer (Chang et al., 2004). Immediate-early genes have unique epigenetic regulation and location in the interphase nucleus. Nucleoporin 153, CTCF, and cohesion are involved in the induction of immediate-early genes. Interestingly, the immediate-early gene FOS is positioned near the nuclear periphery when induced with epidermal growth factor (EGF) in HeLa cells, consistent with nucleoporin 153 determining the spatial positioning of the immediate-early gene (Kadota et al., 2020). We also observed that TPA-induced immediate-early PTGS2 alleles were located at the nuclear periphery of CCD-1070Sk cells (Khan et al., 2017). Another feature of immediate-early gene induction is the requirement for KAT2B-mediated acetylation of the CTD tail domain of RNA polymerase II (Lyons et al., 2020; Schro¨der et al., 2013). Several transcription factors involved in the induction of the immediate-early genes are substrates of kinases activated by the signal transduction pathways. Growth factors and other mitogens engage the RAS-mitogen-activated protein kinase (MAPK) signal transduction pathway which results in a series of kinase activa- tions that lead to immediate-early gene induction. Among the protein kinases involved in this gene programming are the histone H3 kinases, which include the mitogen- and stress-activated protein kinases (MSK1 and MSK2) (Komar and Juszczynski, 2020; Sawicka and Seiser, 2012). MSK1 and MSK2, encoded by the RPS6KA5 and RPS6KA4 genes, respectively, are activated by extracellular-signal-regulated kinase (ERK) 1/2 and p38 (stress-activated protein kinase 2) mitogen activated protein kinase (MAPK) pathways. MSK1/2 phosphorylate nucleosomal histone H3 at S10 (H3S10ph) or S28 (H3S28ph) (termed the nucleosomal response) and transcription factors (CREB, NF-κB subunit p65, ATF1) (Adewumi et al., 2019). Transcription factors associated with the immediate-early gene upstream promoters and enhancers recruit MSK1/ 2, which phosphorylates H3 of nucleosomes in these regions, and other epigenetic modifiers which support transcription initiation (Drobic et al., 2010).
Stimulation of serum-starved quiescent primary human fibroblasts with EGF or TPA results in phosphorylation of H3S10 and H3S28 on separate alleles, and these phosphorylation events parallel the induction of the immediate-early genes (Khan et al., 2017). EGF stimulates the RAS-MAPK pathway through its binding to the EGF receptor, while TPA is a diacylglycerol mimic that activates protein kinase C and the ERK pathway (Ueda et al., 1996). A distinct differential response to these mitogens is that TPA-stimulated primary human fibroblasts had a 20% reduction in nuclear area that was not observed in the EGF-treated cells. In this report, we explored the differential expression of genes in pri- mary human fibroblasts stimulated with EGF or TPA and determined the role of MSK in the induction of the immediate-early genes.

2. Materials and methods

2.1. Cell culture and treatments
CCD-1070Sk (ATCC® CRL-2091™) cells were cultured in MEM (Minimum essential media, Gibco, Life Technologies, Grand Island, NY, USA), supplemented with 10% (vol/vol) fetal bovine serum (FBS, Grand Island, NY, USA), 1% Antibiotic-Antimycotic (Gibco, Grand Island, NY, USA) and were maintained at 37 ◦C in a humidified atmosphere containing 5% CO2. The 80–90% confluent cells were subjected to trypsi- nization and passaged to new culture plates having fresh media. When cells reached a subconfluency level of ~85–90%, they were serum starved for 48 h in MEM media with 0.1% FBS. After serum starvation, the cells were treated with EGF (30 ng/ml, cat no. PHG0311, Invitrogen, MD, USA (now Life Technologies) or TPA (100 nM, Sigma, Saint Louis Missouri, USA) respectively, for indicated time periods. For MSK inhi- bition studies, serum starved cells were pre-treated with 10 µM H89 (Calbiochem, Cat No. 371962–1 M, EMD Millipore Corp., Billerica, MA, USA) for 30 min followed by treatment with EGF (30 ng/ml) or TPA (100 nM) for indicated time periods. In RT-qPCR validation studies, serum staved cells were treated with EGF or TPA as above. Alternatively, serum starved cells were incubated with 10 µM H89 or 100 nM Rp- cAMPS (EMD Millipore Sigma) 30 min before addition of EGF or TPA plus the inhibitor for 60 min. Treatment was arrested by harvesting the cells by centrifugation (2000 rpm/5 min/4 ◦C) and followed by washing the pellets two times with 1XPBS. The experiment was done with three independent biological replicates.

2.2. RNA isolation, reverse transcription PCR
Total RNA was isolated from untreated and treated CCD-1070Sk using RNeasy Mini Kit (QIAGEN) according to manufacturer’s in- structions. Complementary DNA was generated from total RNA (800 ng) using iScript cDNA synthesis kit (BioRad, Hercules CA, USA.) following the manufacturer’s specifications. SsoAdvance universal SYBR green supermiX (BioRad) was used to perform real-time PCR reactions on a CFX96 Touch™ Real-Time PCR Detection System (BioRad). Supple- mentary Table 1 lists the primer sequences used. The levels of gene expression were normalized against the house keeping gene, 18S rRNA. EXperiments were done with three independent biological replicates. All primers were sequenced using the CancerCare Manitoba Research Institute, University of Manitoba, sequencing facility.

2.3. Library preparation and RNA sequencing
Total RNA was quantified by the Qubit 2.0 Fluorometer (Life Tech- nologies) and run on the Bioanalyzer 2100 (Agilent) to analyze the RNA integrity. RNA (1–5 μg) with RIN 7–10 was used for ribosomal RNA depletion using the Ribo-Zero kit (Illumina) according to the manufac- turer’s instructions. Strand specific library construction and sequencing on the 5500 l SOLiD™ System at the University of Manitoba Next Generation Sequencing Platform, Winnipeg, Manitoba was done as previously described (Jahan et al., 2016). For sequencing on the Illu- mina sequencers, library preparation was done using the NEBNext Ultra Directional RNA Library Prep Kit for Illumina (New England Biolabs). The libraries were sequenced at Genome Quebec, Montreal, Quebec and sequenced on the Illumina HiSeq4000 (PE100) and at the Illumina Epigenetics Platform, University of Lethbridge, Lethbridge, Alberta on the NextSeq 500 system. The sequencing experiments were done in two independent biological replicates with each of the sequencing platforms (Supplementary Table 2). Sequencing data files are available from Gene EXpression Omnibus (GEO).

2.4. Bioinformatic analyses
The reads from the NextSeq 500 were mapped to human reference genome hg19 using LifeScope Genomic Analysis Software v2.5.1 (Life Technologies) with 2-mismatch setting and a minimum mapping quality score of 8. More than 88% of reads were mapped. The reads from the HiSeq4000 were mapped to human reference genome hg19 using STAR software default parameters. The RPKM (Reads Per Kilobase Million) values were obtained using HOMER. Raw read counts were obtained from the sequencing platforms separately. Supplementary Table 2 shows the detailed sample ID, inducer, time point, and platform information. DESeq2 was applied for differentially expressed genes among different samples using the raw read counts as inputs (Love et al., 2014). We applied the default filter criteria: 1) Filtered read count’s sum among samples more than 10; 2) Picked genes with adjusted p-value less or equal to 0.05. Volcano plots for different comparison with significant genes were generated. Top 10 GO Terms were also illustrated through barplots. Heatmaps for differentially expressed genes were plotted. Venn diagrams were generated for comparison among different inducers. Dot plots were drawn to show the changes of signature genes before/after induced by EGF/TPA.

3. Results

3.1. Transcriptional program of human fibroblasts CCD-1070Sk70 in response to EGF or TPA
The human fibroblast cell line CCD-1070Sk was cultured under ATCC guidelines followed by 48 h of serum starvation and subsequent treatment with EGF or TPA with and without MSK/PKA inhibitor H89. For RNA sequencing (RNASeq), total cellular RNA was isolated and following removal of ribosomal RNA, single-stranded pair-ended sequencing libraries were prepared. Sequencing was done on various sequencing platforms [SOLiD sequencer (Applied Biosystems), Illumina NextSeq, Illumina HiSeq4000] at sufficient sequencing depth to quantify gene expression (Supplementary Table 2).
Stimulation of primary human fibroblasts (CCD-1070Sk) with EGF initiates a gene expression program, starting with the immediate-early genes. The cells were stimulated for 1, 2, and 4 h. Fig. 1A shows a hi- erarchical map of the changes in the abundance of transcripts during the 4-hour time course. Transcripts from the immediate-early genes such DUSP1 were elevated at 1 h and returned to low levels at 2 h (Fig. 1B). The one-hour time point was then chosen for further study.
Fig. 1. EGF-induced transcriptome in CCD-1070Sk cells. (A) Hierarchical clustering in EGF-induced CCD-1070Sk cells from RNASeq analysis of total cellular RNA from serum starved and EGF induced human fibroblast cells. Red is increased expression; green is reduced expression. EGF stimulated CCD-1070Sk at SS (serum staved) 0 min, 1, 2, and 4-hour time points, respectively. (B) EGF-induction of DUSP1 in CCD-1070Sk cells. Total RNA was isolated and quantified by real-time RT- qPCR. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2A shows the heatmap of the transcript levels from EGF- stimulated versus serum starved cells from four of the siX samples that were sequenced on three different sequencing platforms. The Volcano plot shown in Fig. 2B shows that the transcripts from the immediate- early genes such as DUSP1, DUSP5, ATF3, PTGS2, and NR4A1 were elevated after the one-hour treatment. The top GO terms for the induced genes were nucleus, transcriptional regulators, and genes involved in apoptotic pathways and negative regulators of the ERK signaling pathway (Supplementary Fig. 1A). Fig. 2C shows the heatmap of the transcript levels from TPA-stimulated versus serum starved cells from four samples that were sequenced on different sequencing platforms. As with EGF stimulation, multiple immediate-early genes with GO terms (Supplementary Fig. 1B) such as regulation of transcription and apoptotic processes were induced.
A comparison of the genes induced by EGF or TPA showed more genes were induced with TPA (182) than with EGF (119) (Supplemen- tary Fig. 2). Of these genes, 62 were induced by both mitogens. Among the common genes were DUSP1, NR4A1, and ATF3. For both mitogens, genes involved in shutting down the ERK signaling pathway (DUSP1), transcriptional regulation, and regulation of apoptosis.

3.2. H89 impact on transcriptional program of human fibroblasts CCD- 1070Sk70 in response to EGF or TPA
H89 is a potent inhibitor of MSK1/2 and protein kinase A. Serum- starved fibroblasts were pre-incubated for 30 min with H89 before the addition of mitogens (EGF or TPA). Before proceeding with these studies, we determined the effect of H89 on transcript levels in the serum starved cells. H89 had a profound impact on the expression of several genes as indicated by a reduced level of their transcripts (Supplementary Fig. S3). Forthwith all treatments with serum starved cells were done with H89 / the mitogen.
The preincubation of serum starved cells did not prevent the EGF- induced expression of several immediate-early genes as shown in Fig. 3A and B. Immediate-early genes, EGR2, EGR3, JUNB, and NR4A1, were induced with EGF in the presence of the inhibitor H89. Similarly, H89 did not prevent the TPA-mediated induction of several immediate- early genes (FOSB, FOS, DUSP2, EGR2, EGR3) (Fig. 3C and D). In contrast to the absence of H89, more genes were induced with EGF H89 (185) than with TPA H89 (62) (Fig. 4A). Among these genes, 53 were in common in being induced by either EGF or TPA in the presence of H89. The overlapping genes were used to further to determine gene ontology network using the BiNGO plug in of Cytoscape (Fig. 4B). Fig. 4C is a zoomed in snapshot focusing on the gene enriched biological functions. Some of the most significant and gene enriched biological processes include regulation of metabolic processes, regulation of biosynthetic processes, regulation of transcription (DNA-dependent/ independent), regulation of gene expression.
However, the level of the EGF- or TPA-mediated induction of several immediate-early genes was altered when H89 was present (Fig. 5). A greater number of genes had an altered expression with TPA H89 relative to EGF H89. When EGF was used as the mitogen, EGR1, NIPAL4, FOS, and DUSP5 expression was attenuated when H89 was present. Interestingly, several genes had a greater expression when H89 was present (HES1, ID3, SNAI1). With TPA, 20 genes (e.g., DUSP1, FOS, JUNB) had decreased expression with H89, while 4 genes (e.g., SNAI1) had greater expression in the presence of H89. Together these studies show that the MSK/PKA inhibitor attenuates the expression of several immediate-early genes induced by either EGF or TPA; however, a few mitogen-induced immediate-early gene tran- scripts had increased levels in the presence of H89.

3.3. Validation of H89 affected genes
H89 is a potent inhibitor of MSK1/2 and PKA. In the next series of studies, we did validation studies by RT-qPCR analyses to confirm the effect of H89 on the EGF- or TPA-induced gene and to determine whether the gene transcripts that were impacted by H89 would also be affected by the PKA inhibitor Rp-cAMPS. When doing these studies, we measured the induction of the gene relative to transcript levels in serum starved cells treated with the inhibitor. In agreement with the RNASeq results, EGR1, FOS, HES1, and SNAI1 were induced by EGF (Fig. 6). Further, the induction of EGR1 and FOS was severely diminished in the presence of H89. The EGF induction of HES1 and SNAL1 was enhanced with H89. The PKA inhibitor, Rp-cAMPS, attenuated the induction of the four genes.
In agreement with the RNASeq results, TPA induced the expression of the FOS, DUSP5, and ID3 genes, while H89 attenuated this induction (Fig. 7). However, although we observed that TPA induced TMEM8.8 and SNALI1 transcript levels, H89 attenuated this induction which is in contrast to the RNASeq results (Supplementary Fig. 4). For all these genes, the PKA inhibitor, Rp-cAMPS attenuated the induction by TPA.

4. Discussion

We previously reported that EGF and TPA induced the expression of immediate-early genes PTGS2, FOS, and FOSL1 in human fibroblast CCD-1070Sk cells (Khan et al., 2017). We also observed that TPA, but not EGF, resulted in a change in nuclear structure of serum-starved CCD- 1070Sk cells. Here we used RNASeq to characterize the global response
Fig. 2. EGF- or TPA-induction of CCD-1070Sk cells. (A) The heatmap for EGF 1 h-induced (EGF1) vs. serum starved (SS) is shown. Heatmap shows the expression of EGF1 vs. SS differentially expressed genes from 4 samples from 2 different sequencing platforms. (B) The volcano plot for EGF1 vs. SS. volcano plot shows the comparison of EGF1 vs. SS from 4 samples from 2 different sequencing platforms. The x-axis shows the Log 2-fold change value, the y-axis shows the – Log P value.
Some significant genes are labeled with gene names in the plot. The genes labeled as red are those with P value < 0.05 and absolute log2 fold change > 2; the genes labeled as black are those with P value > 0.05 and absolute log2 fold change <=2; the genes labeled as green are those with P value > 0.05 and absolute log2 fold change > 2; the genes labeled as blue are those with P value < 0.05 and absolute log2 fold change <=2. (C) Heatmap shows the expression of TPA1 vs. SS differentially expressed genes from 4 samples from 2 different sequencing platforms. The volcano plot for TPA1 vs. SS. (D) Volcano plot shows the comparison of
TPA1 vs. SS from 4 samples from 2 different sequencing platforms. The x-axis shows the Log 2-fold change value, the y-axis shows the – Log P value. Some significant genes are labeled with gene names in the plot as described in (B). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) to EGF or TPA stimulation of serum starved cells. TPA, which activates ERK via protein kinase C and RasGRP1, had a greater impact than EGF in inducing immediate-early genes. For both EGF and TPA stimulated cells, the increase in transcripts was from genes involved in transcription, apoptotic and signal transduction processes. However, the increase in expression of processes involved in positive regulation of RNA poly- merase II was more acute in the TPA treated cells. These results are like those obtained with serum induction of quiescent human fibroblasts (Iyer et al., 1999). In this study, Iyer et al reported maximal level of transcripts for immediate-early transcription factors after one hour treatment of serum-starved CCD-1070Sk cells with serum.
With either EGF or TPA stimulation of quiescent CCD-1070Sk cells, there was an increase in MSK1/2-induced H3S10ph and H3S28ph, which we have shown are associated with the regulatory regions of the immediate-early genes. MSK1/2 is a downstream chromatin modifying enzyme of the RAS-MAPK pathway. The activation and recruitment of MSK to an immediate- early gene upstream promoter region initiates a series of events such as phosphorylation of H3S10 or H3S28, binding of
Fig. 3. Effect of H89 on the EGF- or TPA-induction of CCD-1070Sk cells. (A) The heatmap for H89 + EGF 1 h-induced (H89EGF1) vs. H89. Heatmap shows the expression of H89EGF1 vs. H89 differentially expressed genes from 4 samples from 2 different sequencing platforms. (B) The volcano plot for H89EGF1 vs. H89.
Volcano plot shows the comparison of H89EGF1 vs. H89 from 4 samples from 2 different sequencing platforms. The x-axis shows the Log 2-fold change value, the y- axis shows the – Log P value. Some significant genes are labeled with gene names in the plot as described in the legend to Fig. 2. (C) The heatmap for H89 vs. SS. Heatmap shows the expression of H89 vs. SS differentially expressed genes from 6 samples from 3 different sequencing platforms. (D) The volcano plot for H89 vs. SS. Volcano plot shows the comparison of H89 vs. SS from 6 samples from 3 different sequencing platforms. The x-axis shows the Log 2-fold change value, the y-axis shows the – Log P value. Some significant genes are labeled with gene names in the plot as described in the legend to Fig. 2. the 14-3-3ε and ζ proteins to the phosphorylated H3s and recruitment of chromatin remodelers and lysine acetyltransferases (Drobic et al., 2010). Together these events are thought to aid in inducing transcription activity of the immediate-early genes. In support of this model, we observed that the H3S10ph- or H3S28ph-modified chromatin of the immediate-early genes (PTGS2, FOS) was associated with the elongation form of RNA polymerase II. However, in human primary fibroblasts, inhibition of MSK1/2 activity with H89 did not prevent the EGF or TPA induction of several immediate-early genes, suggesting that mechanisms other than those involving MSK are involved in the induction of the genes. This is not to say that MSK does not have a role in the induction of these genes but shows that MSK activity is not necessary for their in- duction. Knockout studies of mouse MSK1/2 reported that the absence of MSK1/2 activity did not have adverse affects on mouse viability, fertility or obvious phenotype (Bertelsen et al., 2011). Further, immediate-early genes, Fos, Egr1, or JunB were induced when mouse MSK1/2 double knockout embryonic fibroblasts were stimulated with TPA or EGF. Together, these observations show that in primary fibro- blasts, MSK activity is dispensable in induction of several immediate- early genes. However, this is not the case of the mitogen induction of immediate-early genes in immortal mouse 10T1/2 cells and Ras-trans- formed mouse fibroblasts in which MSK activity is required (Drobic et al., 2010; P´erez-Cadahía et al., 2011). H89 is a potent inhibitor of MSK and PKA. H89 had a more profound effect on altering the induction of immediate-early genes induced by TPA and those induced by EGF. Further, for the genes we tested that
Fig. 4. Differential expression of genes in CCD-1070Sk cells induced with EGF or TPA in presence of H89. (A) The Venn diagram for H89vsH89EGF1 differentially expressed genes vs. H89 vs H89TPA1 differentially expressed genes. Venn diagram shows the comparison between H89vsH89EGF1 differentially expressed genes and H89vsH89TPA1 differentially expressed genes. (B) The gene ontology network illustrates the complex network of biological processes regulated by the 53 differ- entially expressed genes. The close-up snapshot (C) highlights the biological processes that harbours more genes compared to other processes.
exhibited attenuation with H89 with either EGF- or TPA-induction also had reduced induction with the PKA inhibitor Rp-cAMPS. PKA has been previously reported to be involved in induction of immediate-early genes such as FOS (Huang et al., 1999). This result contrasts to what we observed with mouse 10T1/2 fibroblasts and Ras-transformed mouse fibroblasts in that H89, but not the PKA inhibitor, attenuating the TPA induction of immediate-early genes Jun, Ptgs2, and Fosl1 (Drobic et al., 2010).
Interestingly, we observed that several EGF-induced immediate- early genes (HES1, SNAL1, and ID3) had greater expression with H89. These three genes code for transcriptional repressors. Inhibition of PKA attenuated the EGF induction of HES1 and SNAL1, suggesting that the MSK activity may be involved in moderating the EGF induction of these genes. Loss of Drosophila JIL-1 kinase and its product H3S10 phos- phorylation resulted in up- and down-regulation of genes (Cai et al., 2014). The mechanism involved in the upregulation of genes in the absence of an H3 kinase activity remains unknown. Possible mecha- nisms are other histone modifications being altered (Cai et al., 2014) and/or involvement of other targets of MSK such as transcription factors.
Although MSK activity is not required in immediate-early gene expression in normal cells, these enzymes appear to be an important downstream target of cancer cells that have become addicted to the RAS- MAPK pathway (Meador and Pao, 2015; P´erez-Cadahía et al., 2011). The differential sensitivity of MSK inhibitors in cancer versus normal cells may be of advantage in treating some cancers.
In summary, here we report the immediate-early gene response of serum-starved human primary fibroblasts induced with EGF or TPA in the presence and absence of the MSK/PKA inhibitor H89. The immediate-early gene response is more robust with TPA than it is with EGF, which likely reflects the signaling pathways stimulated by these two agents as well as the effect of these mitogens on nuclear structure.
Fig. 5. Differential expression of genes in H89-treated in CCD-1070Sk cells induced with EGF or TPA. (A) The change of signature genes before/after EGF1 treatment. Dot plot shows the genes induced after 1 h of EGF stimulation and attenuated by H89 with Log 2-fold change > 2. The x-axis shows the – Log 2-fold change value, the y-axis shows the significant genes. Genes attenuated by H89 are shown in red boXes. (B) The change of signature genes before/after TPA1 treatment with/ without H89. Dot plot shows the genes induced after 1 h of TPA stimulation and attenuated by H89 with Log 2-fold change > 2. The x-axis shows the – Log 2-fold change value, the y-axis shows the significant genes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 6. EGF induction of immediate-early genes. Serum starved cells were pretreated (30 min) or not with H89 or Rp-cAMPS prior to EGF stimulation for 60 min. Total RNA was isolated and quantified by real-time RT-qPCR. Fold changes were normalized to 18S RNA and time 0 values for serum starved, H89 or Rp-cAMPS, respectively. Three biological repeats were done, and the errors bars represent S.E.M. as determined by two-way ANOVA. (A,B) FOS, (C,D) EGR1, (E,F) HES1, (G,H) SNAL1. (A,C,E,G) EGF +/- H89; (B,D,F,H) EGF +/- Rp-cAMPS.
Fig. 7. TPA induction of immediate-early genes. Serum starved cells were pretreated (30 min) or not with H89 or Rp-cAMPS prior to TPA stimulation for 60 min. Total RNA was isolated and quantified by real-time RT-qPCR. Fold changes were normalized to 18S RNA and time 0 values for serum starved, H89 or Rp-cAMPS, respectively. Three biological repeats were done, and the errors bars represent S.E.M. as determined by two-way ANOVA. (A,B) FOS, (C,D) DUSP5, (E,F) ID3, (A,C,E) TPA +/- H89; (B,D,F) TPA +/- Rp-cAMPS.
Most induced immediate-early genes were not affected by the presence of H89, demonstrating that MSK and PKA activity were not required for their induction. However, several immediate-early genes required PKA activity. An interesting observation was that H89 but not the PKA in- hibitor Rp-cAMPS increased the EGF induction of genes involved in gene repression. Further studies will be required to understand the mecha- nisms by which MSK may be involved in attenuating the expression of these genes. To further appreciate the mechanisms involved in the in- duction of the immediate-early genes, we need to know more about the regulatory elements (enhancers) involved in controlling the transcrip- tion of these genes. FOS, for example, has multiple enhancers that may be active in different cell types (Kim et al., 2010). A combination of chromatin immunoprecipitation sequencing (H3K27ac, H3K4me3) and transcriptome mapping [e.g., TT-Sequencing (Schwalb et al., 2016)] would be informative. Further investigations are required to learn whether immediate-early genes in general are located at the nuclear periphery (Kadota et al., 2020). MSK activity is required in the expression of immediate-early genes in several cancer cell types. Whether MSK activity is altering transcription factor activity and/or regulatory element chromatin structure differentially in cancer versus normal cells will be important in exploring the use of MSK inhibitors as therapeutic agents.

CRediT authorship contribution statement
Kiran L. Sharma: Formal analysis, Writing – original draft. Shuo Jia: Formal analysis. Tasnim H. Beacon: Formal analysis, Writing – original draft, Writing – review & editing. Ifeoluwa Adewumi: Validation. Camila Lo´pez: Validation. Pingzhao Hu: Writing – review & editing, Supervision. Wayne Xu: Formal analysis. James R. Davie: Conceptu- alization, Writing – original draft, Supervision.

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.

Acknowledgements
This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (RGPIN-2017-05927) and CancerCare Manitoba Foundation (761020318) to JRD.

Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi. org/10.1016/j.gene.2021.145842.

References

Adewumi, I., Lo´pez, C., Davie, J.R., 2019. Mitogen and stress- activated protein kinase regulated gene expression in cancer cells. Adv. Biol. Regul. Elsevier Ltd. 71, 147–155. https://doi.org/10.1016/j.jbior.2018.09.010.
Bertelsen, T., Iversen, L., Riis, J.L., Arthur, J.S.C., Bibby, B.M., Kragballe, K., Johansen, C., 2011. The role of mitogen- and stress-activated protein kinase 1 and 2 in chronic skin inflammation in mice. EXp. Dermatol. 20, 140–145. https://doi.org/ 10.1111/j.1600-0625.2010.01153.X.
Cai, W., Wang, C., Li, Y., Yao, C., Shen, L., Liu, S., Bao, X., Schnable, P.S., Girton, J., Johansen, J., Johansen, K.M., 2014. Genome-wide analysis of regulation of gene expression and H3K9me2 distribution by JIL-1 kinase mediated histone H3S10 phosphorylation in Drosophila. Nucleic Acids Res. 42, 5456–5467. https://doi.org/10.1093/nar/gku173.
Chang, H.Y., Sneddon, J.B., Alizadeh, A.A., Sood, R., West, R.B., Montgomery, K., Chi, J.- T., Van De Rijn, M., Botstein, D., Brown, P.O., 2004. Gene expression signature of fibroblast serum response predicts human cancer progression: similarities between tumors and wounds. PLoS Biol. 2, E7. https://doi.org/10.1371/journal. pbio.0020007.
Drobic, B., P´erez-Cadahía, B., Yu, J., Kung, S.K.P., Davie, J.R., 2010. Promoter chromatin remodeling of immediate-early genes is mediated through H3 phosphorylation at either serine 28 or 10 by the MSK1 multi-protein complex. Nucleic Acids Res. 38, 3196–3208. https://doi.org/10.1093/nar/gkq030.
Fowler, T., Sen, R., Roy, A., 2011. Regulation of primary response genes. Mol. Cell 44, 348–360. https://doi.org/10.1016/j.molcel.2011.09.014.
Huang, Y.T., Lin, J.K., Lee, M.T., 1999. Inhibition of 12-O-tetradecanoylphorbol-13-ac- etate induction of c-fos mRNA by the protein kinase A inhibitor N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinoline sulfonamide. Biochem. Pharmacol. 58, 1639–1647. https://doi.org/10.1016/S0006-2952(99)00238-5.
Iyer, V.R., Eisen, M.B., Ross, D.T., Schuler, G., Moore, T., Lee, J.C.F., Trent, J.M., Staudt, L.M., Hudson, J., Boguski, M.S., Lashkari, D., Shalon, D., Botstein, D., Brown, P.O., Hudson Jr., J., Boguski, M.S., Lashkari, D., Shalon, D., Botstein, D.,
Brown, P.O., 1999. The transcriptional program in the response of human fibroblasts to serum. Science (80-) 283, 83–87. https://doi.org/10.1126/science.283.5398.83.
Jahan, S., Xu, W., He, S., Gonzalez, C., Delcuve, G.P., Davie, J.R., 2016. The chicken erythrocyte epigenome. Epigenetics Chromatin 9, 1–11. https://doi.org/10.1186/ s13072-016-0068-2.
Kadota, S., Ou, J., Shi, Y., Lee, J.T., Sun, J., Yildirim, E., 2020. Nucleoporin 153 links nuclear pore complex to chromatin architecture by mediating CTCF and cohesin MRTX-1257 binding. Nat. Commun. 11 https://doi.org/10.1038/s41467-020-16394-3.
Khan, D.H., Healy, S., He, S., Lichtensztejn, D., Klewes, L., Sharma, K.L., Lau, V., Mai, S., Delcuve, G.P., Davie, J.R., Matera, A.G., 2017. Mitogen-induced distinct epialleles are phosphorylated at either H3S10 or H3S28, depending on H3K27 acetylation. Mol. Biol. Cell 28, 817–824. https://doi.org/10.1091/mbc.e16-08-0618.
Kim, T.-K., Hemberg, M., Gray, J.M., Costa, A.M., Bear, D.M., Wu, J., Harmin, D.A., Laptewicz, M., Barbara-Haley, K., Kuersten, S., Markenscoff-Papadimitriou, E.,
Kuhl, D., Bito, H., Worley, P.F., Kreiman, G., Greenberg, M.E., 2010. Widespread transcription at neuronal activity-regulated enhancers. Nature 465, 182–187. https://doi.org/10.1038/nature09033.
Komar, D., Juszczynski, P., 2020. Rebelled epigenome: histone H3S10 phosphorylation and H3S10 kinases in cancer biology and therapy. Clin. Epigenetics. https://doi.org/ 10.1186/s13148-020-00941-2.
Love, M.I., Huber, W., Anders, S., 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15 https://doi.org/ 10.1186/s13059-014-0550-8.
Lyons, D.E., McMahon, S., Ott, M., 2020. A combinatorial view of old and new RNA polymerase II modifications. Transcription 11, 66–82. https://doi.org/10.1080/ 21541264.2020.1762468.
Meador, C.B., Pao, W., 2015. Old habits die hard: addiction of BRAF-mutant cancer cells to MAP kinase signaling. Cancer Discov 5, 348–350. https://doi.org/10.1158/2159- 8290.CD-15-0221.
P´erez-Cadahía, B., Drobic, B., Espino, P.S., He, S., Mandal, S., Healy, S., Davie, J.R., 2011. Role of MSK1 in the malignant phenotype of Ras-transformed mouse fibroblasts. J. Biol. Chem. 286, 42–49. https://doi.org/10.1074/jbc.M110.156687.
Sawicka, A., Seiser, C., 2012. Histone H3 phosphorylation – a versatile chromatin modification for different occasions. Biochimie 94, 2193–2201. https://doi.org/ 10.1016/j.biochi.2012.04.018.
Schro¨der, S., Herker, E., Itzen, F., He, D., Thomas, S., Gilchrist, D., Kaehlcke, K., Cho, S., Pollard, K., Capra, J., Schno¨lzer, M., Cole, P., Geyer, M., Bruneau, B., Adelman, K., Ott, M., 2013. Acetylation of RNA polymerase II regulates growth-factor-induced gene transcription in mammalian cells. Mol. Cell 52, 314–324. https://doi.org/ 10.1016/j.molcel.2013.10.009.
Schwalb, B., Michel, M., Zacher, B., Frühauf, K., Demel, C., Tresch, A., Gagneur, J., Cramer, P., 2016. TT-seq maps the human transient transcriptome. Science (80-) 352, 1225–1228. https://doi.org/10.1126/science:aad9841.
Ueda, Y., Hirai, S.-I., Osada, S.-I., Suzuki, A., Mizuno, K., Ohno, S., 1996. Protein kinase C δ activates the MEK-ERK pathway in a manner independent of Ras and dependent on Raf. J. Biol. Chem. 271, 23512–23519. https://doi.org/10.1074/jbc.271.38.23512.