Background
Interferons (IFNs) are a group of immunomodulatory proteins made and released by host cells in response to the presence of pathogens such as viruses, bacteria, parasites or tumor cells. IFNs are primarily known for their role in inhibiting viral infections and in stimulating the immune system [
1]. IFN-γ, a type II interferon, is a pleiotropic cytokine involved in the regulation of immune and inflammatory responses, including the activation, growth and differentiation of B-cells, T-cells, natural killer (NK) cells and macrophages. Aberrant IFN-γ expression, associated with autoinflammatory and autoimmune diseases, acts as a pro-inflammatory cytokine [
2]. IFN-γ is produced predominantly by natural killer and natural killer T cells as part of the innate immune response, and by CD4 Th1 and CD8 cytotoxic T lymphocyte effector T cells once antigen-specific immunity develops [
3].
IFN-γ induces the JAK/STAT pathway leading to the activation and binding of transcriptional activators that induce expression of IFN-stimulated genes, notably IRF-1. IRF-1 is expressed at low levels in unstimulated cells and recognizes the ISRE found in most IFN-inducible gene promoters containing repeating short GAAA core sequences [
4,
5]. We have previously shown that CEACAM1 transcription can be induced by IFN-γ via an ISRE in the CEACAM1 promoter [
6‐
8]. CEACAM1 is a member of the immunoglobulin super family of glycoproteins [
9]. It is expressed on the surface of epithelial and activated endothelial cells, as well as on cells from the immune system and plays a role in a variety of cellular processes including cell-cell adhesion, differentiation, apoptosis and immune response [
10].
In humans and rat, CEACAM1 pre-mRNA undergoes extensive AS generating isoforms consisting of an N-terminal and a variable number of multiple extracellular Ig-like domains, a transmembrane domain, and either short (CEACAM1-S) or long (CEACAM1-L) cytoplasmic domains [
11‐
13]. The short cytoplasmic domain isoform, the predominant isoform in epithelial cells, has been shown to bind actin, tropomyosin, calmodulin, and annexin II and is involved in lumen formation [
14‐
16]. In normal breast, it is the predominant isoform, while in breast cancer the S/L ratio is greatly reduced [
17]. The long cytoplasmic domain isoform with its two cytosolic phosphotyrosine residues and immunoreceptor tyrosine-based inhibitory motifs (ITIM), predominant in immune cells, bind SHP-1 when phosphorylated and convey inhibitory activities to CEACAM1-L [
18]. Furthermore, in T lymphocytes CEACAM1-L isoforms are significantly upregulated at the cell surface in response to IL-2 treatment, mediating their cell adhesion to other lymphocytes or tumor cells [
19,
20].
The observation that CEACAM1 functions in recognition and binding between cells with adhesion properties is integral to its signal transduction leading to growth suppressive activities [
21]. Indeed, CEACAM1 is a tumor suppressor gene that is down-regulated in most tumor cells such as colon, breast and prostrate cancer [
21‐
23]. Introduction of CEACAM1-L but not CEACAM1-S cDNA by adenovirus as a gene delivery vehicle reduces tumorigenicity of breast cancer cells [
21]. These findings are consistent with the idea that the expression of CEACAM1 plays a role in keeping cells in a differentiated state, while its down-regulation may be an essential event in the development of epithelial malignancy [
24]. However, the implication of the regulation of its AS generated isoforms is less well understood.
AS is a genomic evolutionary strategy that generates proteome expansion in metazoans [
25,
26].
cis- Regulatory elements known as exonic splicing enhancers and exonic splicing silencers (ESS ) provide the binding platform for splicing regulators to promote or inhibit the use of nearby splice-sites (ss). AS effectors include members of the heterogeneous nuclear ribonucleoprotein (hnRNP) family and serine/arginine-rich (SR) protein families are among the most common splicing regulators [
27,
28]. hnRNPs rapidly associate with nascent transcripts and repress certain AS events leading to exon skipping events [
25]. Three abundant hnRNPs (hnRNP L, hnRNP A1, and hnRNP M) are implicated in many AS events in human and several other eukaryotes and play a role in regulating splicing events by inhibiting the use of splice-sites from the spliceosome [
11]. Previously we showed that hnRNP L and hnRNP A1 interact specifically with exon 7 to control splice-site recognition in the formation of CEACAM1-S mRNA. hnRNP M is important for the AS of CEACAM1-L mRNA [
11].
We recently showed that IFN-γ altered the ratio of splice isoforms of CEACAM1 through the presumable induction of its down-stream effector IRF-1 [
8]. The goal of the present study was to determine whether forced expression of IRF-1 directly plays a role in regulating CEACAM1 AS. Using GFP-driven reporter constructs containing the CEACAM1 genomic promoter and its exons 6-7-8 (heretofore called CAM1 6-7-8) cassettes, we show that AS depends on IRF-1 to couple transcriptional complex communication with the splicing machinery. Transcriptome sequencing analysis shows that IRF-1 can induce the AS of other genes as well, including those involved in membrane regeneration and repair, microtubule assembly, and tumor suppressor functions. We also report extensive changes in transcriptional and AS regulation of cytokines in breast cancer epithelial cells in response to IRF-1 treatment. The AS regulatory activities of IRF-1 demonstrate that the mechanism of the interferon response includes a global change in both gene transcription and AS in breast epithelial cells.
Conclusions
Taken together, our results reveal a novel role for IRF-1 in the cotranscriptional processing of RNA. In this study, we have shown that promoter regulation by IRF-1 determines CEACAM1 AS, specifically to generate the cytoplasmic L-isoform. Effector responses by IRF-1 in the context of AS should reveal novel cellular functions. Our work is consistent with the idea that immune cells reacting to invading areas of a growing tumor generate IRF-1 in the tumor perhaps through release of IFN-γ, which eventually leads to AS to generate CEACAM1 L-isoform expression not only on T cells but in non-inflamed epithelial cells as well. In cases of acute infection this sort of temporary change in phenotype may be advantageous, while a chronic pro-inflammatory phenotype associated with tumor growth may favor further tumor growth.
Methods
Cell culture, reagents, and treatments
MDAMB468 (ATCC® HTB132™) and MCF7 (ATCC® HTB-22) cells were purchased from ATCC and grown in a 5% CO
2 incubator at 37°C in MEM supplemented with 1% Sodium Pyruvate, 1.5% Sodium Bicarbonate, 1 × Non-essential Amino Acids, 1 × Penicillin/Streptomycin/Amphotericin B and 10% heat-inactivated FBS. For induction of IRF-1, mouse IRF-1 was used essentially as described previously except for the following modifications [
29]. Cells were seeded at a density of 0.5 × 10
6 cells in 6-well plates 24 h prior to treatment. Recombinant mouse IRF-1 adenovirus (hereafter referred to as Ad-IRF-1) or the Ad-null empty vector control was added to cultures at PFU estimated at 100 particles per PFU. Cells were incubated for 24 h at 37°C and 5% CO
2 unless otherwise specified. After incubation, RNA and proteins were isolated as described previously [
11]. For Western blots, FACS and immunofluorescent staining identifying CEACAM1 isoforms anti-CEACAM1-L (229; [
54]), T84.1 and 5F4 were used. For CEACAM1 specific antibody, 5F4 was a kind gift of R. Blumberg (Harvard University). Antibodies used for Western blots identifying IRF-1 (ab26109), IRF-2 (C-19, sc-498), GAPDH (FL-335, sc-25778) and p21 (EA10, ab16767) were purchased from Santa Cruz Biotechnology or from Abcam.
RNA isolation semi and quantitative RT-PCR
Total RNA was isolated by the RNeasy mini kit (Qiagen). The RNA was treated with RNase-free DNase set (Qiagen), and RNA (1 μg) was reverse-transcribed in a 20 μl reaction using Maxima First Strand cDNA Synthesis Kit for RT-PCR (K1641, Thermo Scientific) following manufacturer instructions. Reactions were diluted 5-fold and 1/50 of cDNA were used for semiquantitative PCR with gene-specific primers and 2x Taq PCR Premix (Bioland Scientific) for 30 cycles for standard RT-PCR. Primers designed to amplify CEACAM1-3 and CEACAM1-4 isoform were described previously [
7]. Amplification of CEACAM1 mRNA for RT-PCR was performed by using sense exon 6 primers (5’-TTCTGCATTTCGGGAAGACCGGCAG-3’) coupled with antisense 3’ UTR primers (5’-AGCCTGGAGATGCCTATTAG-3’). Primer amplification regions for GFP are indicated in Figure
2C. The products were resolved on 2.25% agarose gels and visualized by staining with SYBR Green I (Invitrogen). Gels were photographed on a GelLogic 200 Imaging System. For real time PCR, 1/50 of cDNA reverse transcription reactions were used to amplify constitutively expressed housekeeping gene β-Actin and CEACAM1 mRNA levels. The reactions were performed on a CFX96 Touch Real-Time PCR Detection System (Bio-Rad) in a 20 μl volume with SsoFast EvaGreen Supermix (Bio-Rad). Primers BP56 (5’- GGTTGCTCTGATAGCAGTAGCCCTGGCATGTTTTCTGCATTTCGGGAAGACCGGCAGCTC-3’) and BP60 with a common sense primer to exon 6 amplify differentially spliced isoforms, CEACAM1-S and CEACAM1-L, respectively [
55]. TaqMan Array Human Cytokine Network (Life Technologies) plates were carried out according to manufacturer instructions. After denaturation for 30 sec at 95°C, 40 cycles were performed (5 sec at 95°C, 5 sec at 60°C). At the end of the reaction, melting curves were generated between 65°C and 95°C, for every 0.5°C. CEACAM1 mRNA and genes identified in our RNA Seq study were calculated by using β-Actin for normalisation. Quantification of gene expression was calculated by using the ΔCT method [
56].
All DNA constructs were generated by standard cloning procedures and confirmed by sequencing. Construction of promoter CEACAM1 WT ISRE:GFP plasmid proceeded by amplifying a genomic DNA fragment corresponding to -1,135 bp relative to the transcription start site (TSS) from MDA-MB-468 cells using
Bgl II and
Hnd III restriction sites cloned into pZsGreen-1 (Clontech). To construct CEACAM1 WT ISRE:CAM1 6-7-8:GFP plasmid, the
Bam HI restriction site in the CEACAM1 promoter was disrupted to allow ligation of our minigene splicing cassette containing genomic CEACAM1 exons 6 through exon 8, described previously [
11], using
Sac II and
Bam HI restriction sites. AS events that favored only L-isoform led to the excitation of GFP fluorescence. To control for external validity and allow for quantitation of transfection efficiency, promoter and splicing reporter experiments were conducted using a second co-transfected promoter CEACAM1 WT ISRE:mCherry plasmid. Mutations for all constructs were introduced using QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent Technologies) following manufacturer’s instructions. To delete both hnRNP L and hnRNP A1 binding sites in exon 7, the sequence AAGCATGCAA was introduced using GeneArt mutagenesis (Life Technologies). This sequence does not associate with any known cellular factor and therefore does not influence AS [
57].
Fluorescence-based electrophoretic mobility shift assay (fEMSA)
The fEMSA was operated using a Light Shift Chemiluminescent EMSA kit (Thermo Scientific) according to the manufacturer’s instructions with the only exception that IRDye fluorescence-labeled oligonucleotides for ISRE WT or ISRE GG->CC were used to visualize DNA-protein binding interactions as described previously [
58]. EBNA was used as a positive control for the assay reagents. Sequence information for each set of oligos is contained in Figure
3A. The DNA-protein binding reaction buffer was composed of 1X binding buffer supplemented with 5% Glycerol, 10 mM MgCl
2, 200 ng Poly (dI-dC) final concentration and 0.3 pmol IRDye-labeled oligonucleotides in a total volume of 10 μl. Recombinant IRF-1 and IRF-2 (ProSpec) each containing a His-tag were used in binding reactions and the reactions were incubated at room temperature for 30 min before separation of complexes on a 6% DNA Retardation Gel (Invitrogen) at 100 V for 60 min. Detection of DNA binding was performed using Licor Odyssey infrared scanning using the following parameters: 700 and 800 nm channels, focus height 3.5 mm, and intensity set to 5.
Flow cytometric analysis
Promoter and splicing reporters were transfected using cells at a density of 1 × 106 cells using Amaxa Cell Line Nucleofector Kit V (Lonza) and 6 μg DNA from each construct. To pre-heated 6-well plates containing growth media, cells were electroporated and immediately incubated for 24 h before flow cytometric analyses using a BD Fortessa (Becton Dickinson, San Jose, CA). GFP and mCherry fluorescence intensities were detected with 488 nm and 561 nm lasers, respectively. Data were analyzed with FlowJo Software version 8.1.1 (Treestar).
RNA immunoprecipitation
We performed RNA immunoprecipitation (RIP) experiments using the Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, Bedford, MA), according to the manufacturer’s instructions. Antibodies to hnRNP L (4D11, sc-32317), hnRNP A1 (4B10, sc-32301), and hnRNP M (3C181, sc-56702) were purchased from Santa Cruz Biotechnology. The co-precipitated RNAs were detected by RT-PCR. Total RNAs (input controls) and isotype controls were assayed simultaneously to demonstrate that the detected signals were from RNAs specifically binding to CEACAM1 mRNA (data not shown). snRNP 70 antibody was used as a negative control for CEACAM1 binding.
Sequences obtained from Illumina’s CASAVA pipeline were aligned to human genome assembly hg19 using Tophat v1.2.0 [
59]. The expression levels of Refseq transcripts were calculated by the number of reads falling into the exon region of each transcript. The expression data were then quantile normalized and log2 transformed. Differentially expressed transcripts were identified by the following criteria: 1) Chi-square p value < 0.05; 2) Log2 ratio > 1 or < -1; 3) at least one sample has coverage > 10, where the coverage was calculated by the expression level of the transcript divided by the transcript length. The up-regulated and down-regulated genes were then subject to functional annotation analysis with DAVID (
http://david.abcc.ncifcrf.gov/). A customized bioinformatics pipeline was developed to detect exon skipping event in the transcriptome. Briefly, for each Refseq transcript with 3 or more exons, the middle exons excluding the first and last exon were kept for further analysis. When exons were shared by multiple transcripts, only one exon was kept. We defined inclusion reads and exclusion reads as described previously [
60]. The inclusion and exclusion reads were counted in both control and IRF-1 treated samples. The inclusion/exclusion ratio for each sample was then calculated. The potential skipped exons in control sample were identified by the following criteria: 1) inclusion reads in control = 0; 2) exclusion reads in control > 1; 3) inclusion reads in Ad-null > =5; 4) exclusion reads in Ad-null =0; 5) Difference of the inclusion/exclusion ratio Ad-null vs. control sample > 8-fold. Similar approach was used to identify skipped exons in the Ad-IRF-1 sample. These combinations of criteria were used to reduce the chance of false positives. The identified exons were then annotated back to Refseq transcripts and compared to differentially expressed transcripts identified above.
Tumor xenografts in NOD/SCID mice
NOD/SCID transgenic mice were purchased from the Jackson Laboratory. MDA-MB-468 tumor cells (1 × 106) were mixed with Matrigel (BD Biosciences) and orthotopically injected into mammary fat pads. The experiment was terminated after 70 days when tumors reached an average size of 150 mm3. Mouse care and experimental procedures were performed under pathogen-free conditions in accordance with established institutional guidance and approved protocols from the Institutional Animal Care and Use Committee of the Beckman Research Institute of City of Hope.
Immunofluorescent staining
Frozen tumor tissue sections were fixated in 2% paraformaldehyde and permeabilized in methanol. After blocking first using Mouse on Mouse (M.O.M.) Blocking Reagent (Vector Lab) and then in PBS containing 10% goat serum for 1 h each, slides were incubated overnight with antibodies: mouse anti-human CEACAM1 ectodomain (5F4, 1:200), rabbit anti-human -L cytoplasmic tail of CEACAM1 (229, 1:200) [
54] or mouse anti-mouse CC1, a kind gift of K. Holmes (University of Colorado). The following day the slides were washed and secondary antibodies were applied for 1 h (goat anti-rabbit, Alexa Fluor 488 labeled and goat anti-mouse, Alexa Fluor 555, both 1:200, Invitrogen) were visualized using an Olympus IX81 Inverted microscope with Q Imaging Retiga 2000R cooled CCD camera with Image Pro Plus version 6.3 imaging software and 40X magnification. For nuclei staining, Hoechst 33342 (blue) was used at final concentration of 1 μg/ml.
Competing interest
The authors declare that they have no competing financial interests.
Authors’ contributions
Experiments, study design and data analysis were performed by KJD. Animal studies were performed by MK. RIP and splicing assays were performed by DG. XW performed bioinformatics analyses. TN performed western blot and cytokine array analyses. CC helped with transient transfection experiments for reporter assays. JHY provided Ad-IRF-1. The manuscript was written by KJD and edited by JES. All authors read and approved the final manuscript.