Genome-wide analysis of vitamin D receptor (VDR) binding sites in THP-1 human monocyte-like cells highlighted the interleukin 8 gene, also known as chemokine CXC motif ligand 8 (CXCL8). CXCL8 is a chemotactic cytokine with important functions during acute inflammation as well as in the context of various cancers. The nine genes of the CXCL cluster and the strong VDR binding site close to the CXCL8 gene are insulated from neighboring genes by CCCTC-binding factor (CTCF) binding sites. Only CXCL8, CXCL6 and CXCL1 are expressed in THP-1 cells, but all three are up-regulated primary 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) target genes. Formaldehyde-assisted isolation of regulatory elements sequencing analysis of the whole CXCL cluster demonstrated 1,25(OH)2D3-dependent chromatin opening exclusively for the VDR binding site. In differentiated THP-1 cells the CXCL8 gene showed a 33-fold higher basal expression, but is together with CXCL6 and CXCL1 still a primary 1,25(OH)2D3 target under the control of the same genomic VDR binding site. In summary, both in undifferentiated and differentiated THP-1 cells the genes CXCL8, CXCL6 and CXCL1 are under the primary control of 1,25(OH)2D3 and its receptor VDR. Our observation provides further evidence for the immune-related functions of vitamin D.
Citation: Ryynänen J, Carlberg C (2013) Primary 1,25-Dihydroxyvitamin D3 Response of the Interleukin 8 Gene Cluster in Human Monocyte- and Macrophage-Like Cells. PLoS ONE 8(10): e78170. https://doi.org/10.1371/journal.pone.0078170
Editor: Sreeram V. Ramagopalan, University of Oxford, United Kingdom
Received: August 2, 2013; Accepted: September 17, 2013; Published: October 21, 2013
Copyright: © 2013 Ryynänen, Carlberg. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the Academy of Finland and the Juselius Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Chemokines are small (8-11 kDa), pro-inflammatory cytokines that are involved in trafficking, activation and proliferation of many cell types, such as myeloid, lymphoid, epidermal and endothelial cells . The over 50 presently known chemokines have been assigned according to the arrangement of their conserved cysteine motifs into the four classes C, CC, CXC and CX3C [2-4]. Chemokine CXC motif ligand (CXCL) 8, also known as interleukin 8, is the first chemokine discovered some 25 years ago  and a prototypical member of the CXC chemokine family. CXCL8 is one of the most potent neutrophil chemo-attractants in acute inflammation , i.e. it is among the first signals to be expressed and released by the various cell types involved in acute inflammation. CXCL8 binds with similar high affinity to the G-protein-coupled receptors CXCR1 and CXCR2 [7,8] and initiates downstream signaling, such as the regulation of fibroblast growth factor 2 and androgen receptor [9,10], suggesting that CXCL8 is also implicated in the control of cellular proliferation, such as in benign prostate hyperplasia and prostate cancer.
The genes encoding for CXCLs 1-8 form together with a variant of CXCL4, CXCL4.1 , a cluster of nine neighboring genes spanning over 350 kb of chromosome 4. Since chemokines are not stored intracellularly but secreted dependent on a stimulus, their effects rely on transcriptional regulation and de novo protein synthesis. The transcriptional regulation of the whole CXCL cluster is largely elusive, but the transcription factors nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and activator protein 1 are known to control CXCL8 and CXCL1 gene transcription [12-14].
The active compound of the vitamin D endocrine system, 1,25-dihydroxyvitamin D3 (1,25(OH)2D3), is not only involved in calcium and phosphate homeostasis and bone mineralization , but there is both epidemiological and pre-clinical evidence that 1,25(OH)2D3 also has anti-proliferative and immuno-modulatory functions [16,17]. In the context of the latter, it had been reported that in immune-stimulated monocytes 1,25(OH)2D3 is able to reduce the interferon γ-mediated up-regulation of the mRNA expression of the cytokines tumor necrosis factor α, interleukin 6 and 1 and of CXCL8 over a time-span of 48 h . In fact, 1,25(OH)2D3 has also been shown in other studies to counteract to pro-inflammatory signal transduction pathways, such as NF-κB signaling , and has specifically shown to inhibit the NF-κB-mediated up-regulation of CXCL8 .
1,25(OH)2D3 is a nuclear hormone that binds directly to the transcription factor vitamin D receptor (VDR) , which is a member of the nuclear receptor superfamily . VDR, like most other transcription factors, competes with the intrinsic repressive nature of chromatin for access to its genomic binding sites [23,24]. Already in the absence of ligand VDR is able to contact genomic DNA and then preferentially forms complexes with co-repressor proteins  and chromatin modifying enzymes, such as histone deacetylases (HDACs) . However, in the presence of ligand VDR interacts with co-activator proteins and histone acetyltransferases . Therefore, interaction with chromatin and its modifying enzymes is a central element in 1,25(OH)2D3 signaling . A direct modulation of transcription by 1,25(OH)2D3 through the interaction of activated VDR with the basal transcriptional machinery is achieved via the specific association of VDR with its genomic binding sites. VDR binding sites, referred to as response elements, are preferentially formed of a direct repeat of two hexameric binding motif spaced by three nucleotides (DR3) [29,30]. Within the last three years the genome-wide binding of VDR has been determined by chromatin immunoprecipitation (ChIP) coupled with massive parallel sequencing (ChIP-seq) in human lymphoblastoid cells (treated for 36 h with 1,25(OH)2D3 ), in human monocyte-like cells (undifferentiated THP-1, stimulated for 40 min with 1,25(OH)2D3 ), in human colorectal cells (LS180, exposed for 180 min with ligand ) and in human hepatic stellate cells (LX2, incubated for 16 h with the 1,25(OH)2D3 analog MC903 ). These four studies revealed 1,600-6,200 specific VDR binding sites, but only a low percentage of them are identical in all investigated cellular models . Moreover, only approximately 30% of these VDR binding sites carry a DR3-type sequence that has a high similarity score with the consensus sequence. This suggests that there are additional modes of VDR binding to its genomic targets .
In earlier studies [32,37,38] we have demonstrated that THP-1 cells represent a well responding and physiologically meaningful model system for the investigation of 1,25(OH)2D3 signaling in the context of innate immunity and cancer. In this study, we investigated the response of CXCL8 and other members of the CXCL cluster in undifferentiated THP-1 cells (monocyte-like cells) and phorbol 12-myristate 13-acetate (PMA)-differentiated THP-1 cells (M2-type macrophage-like cells). We found that in both forms of THP-1 cells the neighboring genes CXCL8, CXCL6 and CXCL1 are primary 1,25(OH)2D3 targets being controlled by the same genomic VDR binding site. This provides further evidence for the immune-related functions of vitamin D.
Material and Methods
The human acute monocytic leukemia cell line THP-1  was grown in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine, 0.1 mg/ml streptomycin and 100 U/ml penicillin and the cells were kept at 37 °C in a humidified 95% air / 5% CO2 incubator. Prior to mRNA or chromatin extraction, undifferentiated THP-1 cells were grown overnight in a density of 500,000 cells/ml in phenol red-free RPMI 1640 medium supplemented with 5% charcoal-stripped FCS. For the differentiation into M2-type macrophage-like cells, THP-1 cells were grown 72 h in normal culture medium supplemented with 20 nM PMA (Sigma-Aldrich). In 1,25(OH)2D3 stimulation experiments, cells were treated with 10 nM 1,25(OH)2D3 (Sigma-Aldrich) or solvent (0.001% ethanol). In HDAC inhibition experiments, cells were stimulated with 300 nM TsA, 2 µM suberoylanilide hydroxamic acid (SAHA), 1 mM valproic acid (VPA, all compounds from Sigma-Aldrich), 100 nM 1,25(OH)2D3 or solvent (0.16% ethanol or 0.02% DMSO).
RNA extraction, cDNA synthesis and PCR
Total RNA extraction, cDNA synthesis and qPCR were performed as described previously . qPCR reactions were performed with the LightCycler® 480 System (Roche) using 400 nM of reverse and forward primers, 2 to 4 µl cDNA or ChIP template and the LightCycler 480 SYBRGreen I Master mix (Roche) or the MaximaTM SYBR Green/ROX qPCR Master mix (Fermentas). Primer-specific temperatures are listed in Tables S1 and S2. Relative mRNA expression levels were determined using the formula 2-(ΔCt), where ΔCt is Ct(target gene) – Ct(reference gene). Like in previous studies on interleukin gene expression in THP-1 cells [37,38], in most experiments RPLP0 was used as a reference gene for normalization. However, in HDAC inhibition experiments, where more global effects on gene expression were anticipated, the target genes were normalized to the three reference genes B2M, GAPDH and HPRT1 as described previously .
ChIP-seq and ChIA-PET data visualization
Publically available CTCF ChIP-seq datasets of the ENCODE consortium  were downloaded from UCSC (http://genome.ucsc.edu/ENCODE) for K562 human monocytic leukemia cells (wgEncodeEH002279), HUVEC human endothelial cells (wgEncodeEH000054) and NHEK human epidermal keratinocytes (wgEncodeEH000063). Our own VDR ChIP-seq (GSE27437) and FAIRE-seq datasets (GSE40075) are available at GEO (www.ncbi.nlm.nih.gov/geo). The Integrative Genomics Viewer (IGV)  was used to visualize ChIP-seq and FAIRE-seq data. The chromatin interaction analysis with paired-end tag sequencing (ChIA-PET) data for CTCF-mediated chromatin loops in K562 cells (wgEncodeEH002075) was visualized using the UCSC genome browser (http://genome.ucsc.edu) .
ChIP was performed as described previously . After 10 min crosslinking, undifferentiated THP-1 cells were collected by centrifugation, while adherent PMA-differentiated THP-1 cells were scraped into Farnham Lysis buffer (0.5% NP-40, 85 mM KCl, protease inhibitors, 5 mM PIPES, pH 8.0) and then pelleted. Immunoprecipitation was carried out by using 1 µg of anti-VDR antibody (sc-1008, Santa Cruz Biotechnologies), CTCF antibody (12-309, Millipore) or non-specific IgG (12-370, Millipore), which were pre-bound to 20 µl Magna ChIP™ Protein A Magnetic Beads (Millipore). Before DNA isolation, samples were reverse cross-linked at 65 °C for 5 h in the presence of proteinase K (Roche) in a final concentration of 100 µg/ml. Equal DNA amounts of chromatin fragments, measured with Quant-iT™ PicoGreen® dsDNA Assay Kit (Invitrogen), were analyzed by qPCR.
The CXCL gene cluster
Formaldehyde-assisted isolation of regulatory elements sequencing (FAIRE-seq) is a method that allows the identification of chromatin sites devoid of nucleosomes, roughly translating to the genome-wide localization of chromatin regions that are accessible to transcription factors, such as VDR, at a given time and condition [48,49]. In this study, we used a FAIRE-seq dataset obtained from THP-1 human monocytic leukemia cells [50,51] and aligned the resulting peaks with the VDR ChIP-seq dataset from the same cell line . Interestingly, a chromatin region spanning over 180 kb (from 45 kb upstream of the CXCL8 gene to 9 kb downstream of the CXCL1 gene, underlined in the top lane of Figure 1A) displayed a higher rate of open chromatin, since it showed stronger FAIRE signals than its up- and downstream flanking regions. VDR ChIP-seq analysis in THP-1 cells  of the same genomic region around the CYCL cluster highlighted a prominent, 1,25(OH)2D3-inducible VDR binding site 22 kb downstream of the transcription start site (TSS) of the CXCL8 gene (Figure 1A).
A. The IGV browser was used to show the peak tracks of FAIRE-seq data from THP-1 cells  (stimulated for 20 min with ethanol, turquoise) and VDR ChIP-seq data from THP-1 cells  (unstimulated (-) and treated for 40 min with 1,25(OH)2D3 (+), red). The gene structures are shown in blue and the 9 genes of the CXCL gene cluster are underlayed in grey. The THP-1 data were compared with CTCF ChIP-seq data from the ENCODE cell lines K562, HUVEC and NHEK  (blue) and CTCF ChIA-PET data  in track view (dark blue) and in looping view (grey horizontal lines). Six conserved CTCF sites were highlighted. B. ChIP-qPCR was performed with chromatin samples obtained from unstimulated THP-1 cells to determine CTCF (blue) and unspecific IgG (grey) binding at the six genomic regions, which were suggested by data obtained in K562 cells (see panel A). Columns represent the means of at least three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed to determine the significance CTCF association in reference to a control region from chromosome 6 (* p < 0.05; ** p < 0.01; *** p < 0.001).
The sequence-specific transcription factor CTCF is known as a chromatin organizer, which links chromosomal domains . Therefore, when neighboring genes are co-regulated, they should to be part of the same chromosomal domain and not be separated by insulating CTCF binding sites. Since CTCF binding sites are highly conserved between tissues and cell lines , we used the publically available ChIP-seq datasets of the ENCODE consortium  and displayed CTCF binding sites from K562 human monocytic leukemia cells, HUVEC human endothelial cells and NHEK human epidermal keratinocytes over the whole CXCL cluster and its upstream and downstream flanking genes (Figure 1A). A genome-wide map of the 3-dimensional interactions of CTCF in K562 cells, as obtained by ChIA-PET assays , was displayed by using the UCSC genome browser (Figure 1A). It demonstrated that the chromosomal domain containing the whole CXCL gene cluster, but not any other genes, spans from a region upstream of CTCF binding site 2 to site 5, i.e. over nearly 500 kb. Moreover, a number of loops connect CTCF site 1 to upstream and CTCF site 6 a to downstream CTCF sites, respectively, i.e. the genes that are located in these flanking regions seem to be part of different chromosomal domains than that of the isolated CXCL gene cluster.
In order to confirm that the CTCF binding sites, which are suggested by ENCODE data, are also used in our THP-1 cell model, we performed ChIP-qPCR with chromatin samples from non-stimulated, undifferentiated THP-1 cells (Figure 1B). In comparison to a negative control region from chromosome 6, we found to all six genomic regions CTCF binding. This means that also in THP-1 cells these conserved CTCF sites are occupied with protein.
In summary, the cluster of nine CXCL genes carries a prominent VDR binding site close to the CXCL8 gene and is flanked by conserved CTCF binding sites, which define the borders of the chromosomal domain of the gene cluster.
1,25(OH)2D3 response of CXCL cluster genes in undifferentiated THP-1 cells
In order to get an overview on the relative basal expression of the members of the CXCL cluster and their upstream and downstream flanking genes, we performed qPCR in non-stimulated, undifferentiated THP-1 cells (Figure 2A). Within the CXCL cluster we could detect the expression of only CXCL8, CXCL6 and CXCL1: CXCL8 is 28- and 18-times higher expressed than CXCL6 and CXCL1, respectively. In addition, from the upstream flanking genes of the CXCL cluster the albumin (ALB) gene and from the downstream flanking the genes methylenetetrahydrofolate dehydrogenase (NADP+-dependent) 2-like (MTHFD2L) and amphiregulin (AREG) are expressed in undifferentiated THP-1 cells. Next we stimulated the cells with 1,25(OH)2D3 and performed qPCR for the six expressed genes, in order to evaluate their possible primary response to the VDR ligand. Interestingly, the detailed time courses indicated that CXCL8 (Figure 2B), CXCL6 (Figure 2C) and CXCL1 (Figure 2D) are already significantly up-regulated 1 h after onset of stimulation with 1,25(OH)2D3 and reach after 8 h an induction of 9.1-fold for CXCL8, 3.7-fold for both CXCL6 and CXCL1, respectively. In contrast, the flanking genes ALB, MTHFD2L and AREG display no significant response to 1,25(OH)2D3 (data not shown).
With samples obtained from THP-1 cells qPCR was performed to determine the basal expression, relative to the housekeeping gene RPLP0, of the nine genes of the CXCL gene cluster and each four flanking genes (A) and the change of expression of CXCL8 (B), CXCL6 (C) and CXCL1 (D) in response to incubation with 10 nM 1,25(OH)2D3 over a time period of 8 h. Columns (A) and data points (B-D) represent the means of at least three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed to determine the significance of the mRNA induction by the stimuli (* p < 0.05; ** p < 0.01; *** p < 0.001).
Taken together, in undifferentiated THP-1 cells only the CXCL cluster genes CXCL8, CXCL6 and CXCL1 are expressed, but all three are primary 1,25(OH)2D3 targets. From the genes flanking the CXCL cluster only ALB, MTHFD2L and AREG are expressed, but none of them responds to 1,25(OH)2D3 stimulation.
Open chromatin within the CXCL cluster
Open chromatin is in general more sensitive to HDAC inhibitors than closed chromatin. Therefore, we assessed CXCL8, CXCL6 and CXCL1 gene expression after inhibition of HDACs by TsA, SAHA and VPA for 2.5 and 24 h alone and in combination with 1,25(OH)2D3 (Figure S1). After short-term HDAC inhibitor treatment all three genes were down-regulated: CXCL8 by VPA, CXCL6 by SAHA and CXCL1 by TsA. In contrast, after 24 h CXCL8 was up-regulated by SAHA, CXCL6 even by both SAHA and VPA, while CXCL1 showed no response. The 2.5 h treatment with 1,25(OH)2D3 resulted for all three genes in an approximately 2-fold up-regulation, which is consistent with our time course data (Figure 2). Furthermore, the 24 h time point indicated a prominent long-term stimulation of all three genes by 1,25(OH)2D3: 32-fold for CXCL8, 17-fold for CXCL6 and 14-fold for CXCL1. Consistent with our previous findings , at short-term treatment (2.5 h) with 1,25(OH)2D3 together with TsA, SAHA or VPA the HDAC inhibitors dominated over the VDR ligand. At long-term double treatment (24 h), TsA and SAHA significantly reduced the strong 1,25(OH)2D3 up-regulation of CXCL8 and SAHA and VPA that of CXCL1. However, the HDAC inhibitors had no significant effect on the 1,25(OH)2D3 response of the CXCL6 gene.
The FAIRE-seq pattern of the genomic region around the genes CXCL8, CXCL6 and CXCL1 suggests that a treatment with 1,25(OH)2D3 has no global effect on the number or intensity of sites of open chromatin in THP-1 cells (Figure 1A and data not shown). However, we observed at the VDR binding site close to the CXCL8 gene a significant, 1,25(OH)2D3-dependent opening of chromatin in a FAIRE-seq time course experiment with measurements every 20 min over a time period of 120 min (Figure 3A). In order to confirm VDR binding to this site, we performed ChIP-qPCR with chromatin samples from THP-1 cells that had been treated for 0, 1 and 2 h with 1,25(OH)2D3 (Figure 3B). In comparison to a negative control region from chromosome 6, we observed already in the absence of ligand VDR binding to the site, which significantly increased by the addition of 1,25(OH)2D3. From previous studies [32,50,51] we know that VDR binding sites at regions of 1,25(OH)2D3-dependent chromatin opening have genome-wide the highest rate of DR3-type response elements (66%) below VDR ChIP-seq summits. Consistent with this, the VDR binding site close to the CXCL8 gene also contained a sequence with a high similarity score to a DR3-type response element (Figure 3A).
A. The IGV browser was used to display the genomic region around the CXCL8 gene. The peak tracks show VDR ChIP-seq data (red ) and FAIRE-seq data (grey for the ethanol-treated control, turquoise for the samples treated with 1,25(OH)2D3 for indicated times ), both from THP-1 cells. The gene structures are shown in blue. The sequence of a DR3-type VDR binding site below the summit of the VDR ChIP-seq peaks is indicated. B. ChIP-qPCR was performed with chromatin samples obtained from THP-1 cells to determine VDR association (red) and unspecific IgG binding (grey) at the VDR binding sites close to the CXCL8 gene and a negative control region of chromosome 6. Cells were stimulated for 0, 1 and 2 h with 10 nM 1,25(OH)2D3 and chromatin was extracted. Columns represent the means of at least three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed to determine the significance of 1,25(OH)2D3-induced VDR association in reference to untreated cells (** p < 0.01).
In summary, the genes CXCL8, CXCL6 and CXCL1 are sensitive to HDAC inhibitor treatment, which also modulates their response to 1,25(OH)2D3. The VDR binding site close to the CXCL8 gene co-locates with a region of 1,25(OH)2D3-sensitve open chromatin and carries a DR3-type response element.
1,25(OH)2D3 response of CXCL cluster genes in PMA-differentiated THP-1 cells
The phorbol ester PMA is known to differentiate in suspension growing THP-1 cells into adherent M2-type macrophage-like cells . In such PMA-differentiated THP-1 cells we used qPCR to compare the basal expression of the genes of the CXCL cluster and their flanking genes (Figure 4A). In addition to the genes CXCL8, CXCL6, CXCL1, ALB, MTHFD2L and AREG, which are already expressed in undifferentiated THP-1 cells (Figure 2), we found the expression of CXCL7 and CXCL3. Also in differentiated THP-1 cells CXCL8 displays the highest expression within the investigated genomic region and showed 443-fold higher mRNA levels than CXCL6, 114-fold higher than CXCL1, 48-fold more than CXCL7 and a 67-fold excess compared to CXCL3. Moreover, compared to undifferentiated cells, in PMA-differentiated THP-1 cells CXCL8 is 33-times higher expressed. Detailed 8 h time course experiments in PMA-differentiated THP-1 cells showed that CXCL8 (Figure 4B), CXCL6 (Figure 4C) and CXCL1 (Figure 4D) are primary 1,25(OH)2D3 target genes also in this cellular model. However, in these macrophage-like cells all three CXCL genes are less inducible than in undifferentiated THP-1 (monocyte-like) cells: even after 8 h stimulation with 1,25(OH)2D3 the induction of CXCL8 is only 1.9-fold, that of CXCL6 is 3.3-fold and and that of CXCL1 is 3.0-fold. Furthermore, for all three genes a significant induction by 1,25(OH)2D3 was detected only after 2.5 to 3.5 h stimulation, i.e. clearly delayed compared to undifferentiated THP-1 cells. For comparison, in the same time course experiments the genes CXCL7 and CXCL3 showed no significant response to 1,25(OH)2D3 (Figure S2). Moreover, also in PMA-differentiated THP-1 cells the flanking genes ALB, MTHFD2L and AREG do not shown any early response to treatment with 1,25(OH)2D3 (data not shown).
With samples obtained from PMA-differentiated THP-1 cells qPCR was performed to determine the basal expression, relative to the housekeeping gene RPLP0, of the nine genes of the CXCL gene cluster and each four flanking genes (A) and the change of expression of CXCL8 (B), CXCL6 (C) and CXCL1 (D) in response to incubation with 10 nM 1,25(OH)2D3 over a time period of 8 h. Columns (A) and data points (B-D) represent the means of at least three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed to determine the significance of the mRNA induction by 1,25(OH)2D3 (* p < 0.05; ** p < 0.01; *** p < 0.001).
Taken together, in differentiated THP-1 cells the CXCL cluster genes CXCL8, CXCL6 and CXCL1 are higher expressed than in undifferentiated cells. In addition, CXCL7 and CXCL3 expression is detected. However, also in differentiated cells CXCL8, CXCL6 and CXCL1 are the only primary 1,25(OH)2D3 targets within the CXCL cluster, but their inducibility by 1,25(OH)2D3 is reduced and delayed.
VDR binding in PMA-differentiated THP-1 cells
In order to investigate, whether a differentiation of THP-1 cells into macrophage-like cells modulates the VDR binding to the CXCL gene cluster, we performed ChIP-seq for VDR in PMA-differentiated THP-1 cells. In differentiated THP-1 cells we found VDR binding at the same location than in undifferentiated cells (Figure 5A). Moreover, by ChIP-qPCR in PMA-differentiated THP-1 cells we could confirm a ligand-dependent binding of VDR to this site (Figure 5B). Furthermore, we could not detect any additional significant VDR binding site within 3 Mb distance to the CXCL cluster, when the THP-1 cells were differentiated into macrophage-like cells (Figure 5A and data not shown).
A. The IGV browser was used to display the genomic region +/-150 kb around the VDR peak close to the CXCL8 gene. The peak tracks show VDR ChIP-seq data obtained from undifferentiated THP-1 cells (red ) and from PMA-differentiated THP-1 cells (green). The gene structures are shown in blue. B. ChIP-qPCR was performed with chromatin samples obtained from PMA-differentiated THP-1 cells to determine VDR association (red) and unspecific IgG binding (grey) at the VDR binding site and a negative control region of chromosome 6. Cells were stimulated for 0, 1 and 2 h with 10 nM 1,25(OH)2D3 and chromatin was extracted. Columns represent the means of at least three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed but could not determine significant 1,25(OH)2D3-induced VDR association in reference to untreated cells.
In summary, in PMA-differentiated THP-1 cells the CXCL cluster is controlled by the same VDR binding site than in undifferentiated cells.
VDR ChIP-seq and microarray assays performed in undifferentiated THP-1 cells  indicated that the CXCL8 gene may be a target of 1,25(OH)2D3 and its receptor VDR. Therefore, we investigated in this study the 1,25(OH)2D3 response of the whole CXCL gene cluster both in undifferentiated and PMA-differentiated THP-1 cells. We were able to confirm the primary response of CXCL8 to 1,25(OH)2D3 in undifferentiated THP-1 cells and found the neighboring genes CXCL6 and CXCL1 to be primary VDR targets as well. In differentiated THP-1 cells the same three genes are also 1,25(OH)2D3 targets but, while they respond in undifferentiated cells already within 1 h after onset of stimulation, in differentiated cells their response was delayed by 1.5 to 2.5 h. Moreover, the prominent induction of CXCL8 gene expression in undifferentiated cells is, dependent on the time of stimulation, 2- to 5-fold reduced in differentiated cells. However, the reduced responsiveness of CXCL8 to 1,25(OH)2D3 in differentiated cells coincides with a 33-fold higher basal expression, so that in the latter cell type a stimulation with 1,25(OH)2D3 induces even a higher number of de novo synthesized CXCL8 mRNA molecules than in undifferentiated cells. The same applies for the genes CXCL6 and CXCL1, which showed an up to 2-times reduced inducibility by 1,25(OH)2D3, when THP-1 cells differentiate into macrophage-like cells, but increased their basal expression more than 2-fold.
VDR ChIP-seq analysis in undifferentiated and PMA-differentiated THP-1 cells suggests that the CXCL gene cluster is controlled by a single VDR binding site close to the CXCL8 gene. The location of conserved, insulating CTCF binding sites suggest that the VDR binding site and all nine CXCL genes are located within the same chromosomal domain. FAIRE-seq data suggest that in undifferentiated THP-1 cells the genomic region around the genes CXCL8, CXCL6 and CXCL1 is far more accessible than the remaining CXCL cluster. Therefore, it is surprising that CXCL4.1 gene, which is located between CXCL6 and CXCL1, is not expressed in these cells. However, in tissues where CXCL4.1 is expressed, it should be a VDR target gene. In contrast, although the genes CXCL7 and CXCL3 are expressed in differentiated THP-1 cells, they do not respond to stimulation with 1,25(OH)2D3. This suggests that PMA-differentiated THP-1 cells may use the CTCF sites 3 or 4, in order to loop to CTCF cite 2 (see Figure 1A), i.e. that shorter DNA loops may be formed in differentiated cells than in undifferentiated cells.
Nevertheless, the VDR binding of the CXCL cluster belongs to a group of 165 genome-wide locations , for which a stimulation with 1,25(OH)2D3 results in a prominent opening of chromatin at the locus of VDR binding. Within this subset of VDR binding sites 66% carry a DR3-type response element within the sequence below the respective VDR peak summits . This is a more than 2-times higher rate than the 31.7% reported for all genomic VDR binding sites in undifferentiated THP-1 cells . Accordingly, we found below the summit of the VDR peak close to the CXCL8 gene also a DR3-type response element. This suggests that the VDR binding site of the CXCL cluster can be distinguished from the majority of genomic locations of the VDR. We speculate that this site may represent a preferred contact point of the receptor with the genome, which may have been evolutionary selected.
The facts that we observed i) a VDR binding site close to the CXCL8 gene, ii) 1,25(OH)2D3-dependent chromatin opening at the latter site, iii) a DR3-type response element at this site indicating direct DNA binding of the VDR and iv) mRNA up-regulation of CXCL8, CXCL6 and CXCL1 suggest that the three genes are classical, up-regulated primary targets of 1,25(OH)2D3. This conclusion seems to contradict previous reports that 1,25(OH)2D3 represses CXCL8 expression [18,20,53]. However, in these studies CXCL8 gene expression [18,53] or CXCL8 promoter activity  had been stimulated by the cytokines interferon-γ and tumor necrosis factor α, respectively, and by lipopolysaccharide, i.e. by stimuli for transcription factors, such as NF-κB, that strongly up-regulate the CXCL8 gene. Moreover, two of these studies used primary monocytes [18,53] and one a melanoma cell line , i.e. cellular systems different from THP-1 cells. In contrast, a recent study in THP-1 cells, in which the cells were stimulated with a potent synthetic VDR ligand, is consistent with our finding that CXCL8 gene expression is up-regulated . Since VDR is known to negatively interfere with the activity of NF-κB  and other mediators of pro-inflammatory signaling pathways , the observations of Harant et al. , Di Rosa et al.  and Giulietti et al.  most likely represent repressing effects on a transcription factor activating CXCL8 gene expression rather than any primary effect of VDR on the gene’s activity.
The above discussed points suggest that 1,25(OH)2D3 may have a dual effect on CXCL gene expression: a specific primary up-regulation via direct binding of VDR to the CXCL cluster locus and a more global secondary effect, by which all those genes are repressed that are early responding targets of NF-κB. Both effects depend on chromatin organization, i.e. the differentiation status of the cells, and external signals, such as paracrine effects of NF-κB-stimulating cytokines. In general, the actions of 1,25(OH)2D3 are summarized to be anti-inflammatory [56,57], to which the here reported up-regulation of three pro-inflammatory chemokines seem to be a contradiction. However, when the complex time- and signal-dependent inflammatory reaction is separated into its individual phases, it becomes clear that 1,25(OH)2D3 has a modulatory effect on all of them. This implies that in the early phase of inflammation, to which the up-regulation of chemo-attractant CXCL chemokines belongs, 1,25(OH)2D3 has a supporting role, while in the later phase a possible overreaction of the inflammatory response is controlled by 1,25(OH)2D3 via the repression of NF-κB. The potentiation of CXCL8 expression by 1,25(OH)2D3 will lead to an initially more pronounced inflammatory reaction, which, dependent on the type of cancer, will either have an supporting or disadvantageous effect on cell survival . In this way, our observation of the fast up-regulation of CXCL8, CXCL6 and CXCL1 by 1,25(OH)2D3 provides an additional aspect to the effects of the nuclear hormone on the immune response with impact on cancer immunology.
In conclusion, we found that both in undifferentiated and in PMA-differentiated THP-1 cells the genes CXCL8, CXCL6 and CXCL1 are primary targets of 1,25(OH)2D3 and its receptor VDR. Our observation implies a more differential view on the modulation of the inflammatory reaction by 1,25(OH)2D3 and provides further evidence for the impact of vitamin D in supporting the immune system in its fight against both microbes and cancer.
Short- and long-term effects of HDAC inhibition on CXCL genes in undifferentiated THP-1 cells. In THP-1 cells qPCR was performed to determine the relative changes of mRNA expression of the genes CXCL8, CXCL6 and CXCL1 in response to incubation with 100 nM 1,25(OH)2D3 (1,25D), 300 nM TsA, 3 µM SAHA and 1 mM VPA, alone or in combination, for 2.5 and 24 h. The data points represent the means of three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed to determine the significance of the mRNA induction by the stimuli and the repression of the 1,25(OH)2D3 stimulation by HDAC inhibitors (* p < 0.05; ** p < 0.01; *** p < 0.001).
CXCL7 and CXCL3 are no 1,25(OH)2D3 target genes cluster in PMA-differentiated THP-1 cells. With samples obtained from PMA-differentiated THP-1 cells qPCR was performed to determine the change of expression of CXCL7 (A) and CXCL3 (B) in response to incubation with 10 nM 1,25(OH)2D3 over a time period of 8 h. Data points represent the means of at least three independent experiments and the bars indicate standard deviations. Two-tailed Student’s t-tests were performed to determine the significance of the mRNA induction by the stimuli (* p < 0.05).
Reverse transcription qPCR primers.
The authors thank Dr. Sami Heikkinen for discussions and Dr. Sabine Seuter for cDNA samples.
Conceived and designed the experiments: JR CC. Performed the experiments: JR. Analyzed the data: JR CC. Wrote the manuscript: CC.
- 1. Homey B, Müller A, Zlotnik A (2002) Chemokines: agents for the immunotherapy of cancer? Nat Rev Immunol 2: 175-184. doi:10.1038/nri748. PubMed: 11913068.
- 2. Fernandez EJ, Lolis E (2002) Structure, function, and inhibition of chemokines. Annu Rev Pharmacol Toxicol 42: 469-499. doi:10.1146/annurev.pharmtox.42.091901.115838. PubMed: 11807180.
- 3. Le Y, Zhou Y, Iribarren P, Wang J (2004) Chemokines and chemokine receptors: their manifold roles in homeostasis and disease. Cell Mol Immunol 1: 95-104. PubMed: 16212895.
- 4. Bacon K, Baggiolini M, Broxmeyer H, Horuk R, Lindley I et al. (2002) Chemokine/chemokine receptor nomenclature. J Interferon Cytokine Res 22: 1067-1068. doi:10.1089/107999002760624305. PubMed: 12433287.
- 5. Lindley I, Aschauer H, Seifert JM, Lam C, Brunowsky W et al. (1988) Synthesis and expression in Escherichia coli of the gene encoding monocyte-derived neutrophil-activating factor: biological equivalence between natural and recombinant neutrophil-activating factor. Proc Natl Acad Sci U S A 85: 9199-9203. doi:10.1073/pnas.85.23.9199. PubMed: 3057503.
- 6. Vandercappellen J, Van Damme J, Struyf S (2008) The role of CXC chemokines and their receptors in cancer. Cancer Lett 267: 226-244. doi:10.1016/j.canlet.2008.04.050. PubMed: 18579287.
- 7. Nasser MW, Raghuwanshi SK, Grant DJ, Jala VR, Rajarathnam K et al. (2009) Differential activation and regulation of CXCR1 and CXCR2 by CXCL8 monomer and dimer. J Immunol 183: 3425-3432. doi:10.4049/jimmunol.0900305. PubMed: 19667085.
- 8. Singh S, Singh AP, Sharma B, Owen LB, Singh RK (2010) CXCL8 and its cognate receptors in melanoma progression and metastasis. Future Oncol 6: 111-116. doi:10.2217/fon.09.128. PubMed: 20021212.
- 9. Giri D, Ittmann M (2001) Interleukin-8 is a paracrine inducer of fibroblast growth factor 2, a stromal and epithelial growth factor in benign prostatic hyperplasia. Am J Pathol 159: 139-147. doi:10.1016/S0002-9440(10)61681-1. PubMed: 11438462.
- 10. Seaton A, Scullin P, Maxwell PJ, Wilson C, Pettigrew J et al. (2008) Interleukin-8 signaling promotes androgen-independent proliferation of prostate cancer cells via induction of androgen receptor expression and activation. Carcinogenesis 29: 1148-1156. doi:10.1093/carcin/bgn109. PubMed: 18487223.
- 11. Vandercappellen J, Van Damme J, Struyf S (2011) The role of the CXC chemokines platelet factor-4 (CXCL4/PF-4) and its variant (CXCL4L1/PF-4var) in inflammation, angiogenesis and cancer. Cytokine Growth Factor Rev 22: 1-18. doi:10.1016/j.cytogfr.2010.10.011. PubMed: 21111666.
- 12. Amiri KI, Ha HC, Smulson ME, Richmond A (2006) Differential regulation of CXC ligand 1 transcription in melanoma cell lines by poly(ADP-ribose) polymerase-1. Oncogene 25: 7714-7722. doi:10.1038/sj.onc.1209751. PubMed: 16799643.
- 13. Wolf JS, Chen Z, Dong G, Sunwoo JB, Bancroft CC et al. (2001) IL (interleukin)-1α promotes nuclear factor-kappaB and AP-1-induced IL-8 expression, cell survival, and proliferation in head and neck squamous cell carcinomas. Clin Cancer Res 7: 1812-1820. PubMed: 11410524.
- 14. Hoffmann E, Dittrich-Breiholz O, Holtmann H, Kracht M (2002) Multiple control of interleukin-8 gene expression. J Leukoc Biol 72: 847-855. PubMed: 12429706.
- 15. DeLuca HF (2004) Overview of general physiologic features and functions of vitamin D. Am J Clin Nutr 80: 1689S-1696S. PubMed: 15585789.
- 16. Ingraham BA, Bragdon B, Nohe A (2008) Molecular basis of the potential of vitamin D to prevent cancer. Curr Med Res Opin 24: 139-149. doi:10.1185/030079908X253519. PubMed: 18034918.
- 17. Verstuyf A, Carmeliet G, Bouillon R, Mathieu C (2010) Vitamin D: a pleiotropic hormone. Kidney Int 78: 140-145. doi:10.1038/ki.2010.17. PubMed: 20182414.
- 18. Giulietti A, van Etten E, Overbergh L, Stoffels K, Bouillon R et al. (2007) Monocytes from type 2 diabetic patients have a pro-inflammatory profile. 1,25-Dihydroxyvitamin D3 works as anti-inflammatory. Diabetes Res Clin Pract 77: 47-57. doi:10.1016/j.diabres.2006.10.007. PubMed: 17112620.
- 19. Yu XP, Bellido T, Manolagas SC (1995) Down-regulation of NF-kappa B protein levels in activated human lymphocytes by 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci U S A 92: 10990-10994. doi:10.1073/pnas.92.24.10990. PubMed: 7479923.
- 20. Harant H, Andrew PJ, Reddy GS, Foglar E, Lindley IJ (1997) 1α,25-dihydroxyvitamin D3 and a variety of its natural metabolites transcriptionally repress nuclear-factor-kappaB-mediated interleukin-8 gene expression. Eur J Biochem 250: 63-71. doi:10.1111/j.1432-1033.1997.00063.x. PubMed: 9431991.
- 21. Carlberg C, Molnar F (2012) Current status of vitamin D signaling and its therapeutic applications. Curr Top Med Chem 12: 528-547. doi:10.2174/156802612799436623. PubMed: 22242854.
- 22. Perissi V, Rosenfeld MG (2005) Controlling nuclear receptors: the circular logic of cofactor cycles. Nat Rev Mol Cell Biol 6: 542-554. doi:10.1038/nrm1680. PubMed: 15957004.
- 23. Razin A (1998) CpG methylation, chromatin structure and gene silencing-a three-way connection. EMBO J 17: 4905-4908. doi:10.1093/emboj/17.17.4905. PubMed: 9724627.
- 24. Eberharter A, Becker PB (2002) Histone acetylation: a switch between repressive and permissive chromatin. Second in review series on chromatin dynamics. EMBO Rep 3: 224-229. doi:10.1093/embo-reports/kvf053. PubMed: 11882541.
- 25. Polly P, Herdick M, Moehren U, Baniahmad A, Heinzel T et al. (2000) VDR-Alien: a novel, DNA-selective vitamin D3 receptor-corepressor partnership. FASEB J 14: 1455-1463. doi:10.1096/fj.14.10.1455. PubMed: 10877839.
- 26. Malinen M, Saramäki A, Ropponen A, Degenhardt T, Väisänen S et al. (2008) Distinct HDACs regulate the transcriptional response of human cyclin-dependent kinase inhibitor genes to trichostatin A and 1α,25-dihydroxyvitamin D3. Nucleic Acids Res 36: 121-132. doi:10.1093/nar/gkn531. PubMed: 17999998.
- 27. Gronemeyer H, Gustafsson JA, Laudet V (2004) Principles for modulation of the nuclear receptor superfamily. Nat Rev Drug Discov 3: 950-964. doi:10.1038/nrd1551. PubMed: 15520817.
- 28. Carlberg C, Seuter S (2010) Dynamics of nuclear receptor target gene regulation. Chromosoma 119: 479-484. doi:10.1007/s00412-010-0283-8. PubMed: 20625907.
- 29. Pike JW, Meyer MB, Watanuki M, Kim S, Zella LA et al. (2007) Perspectives on mechanisms of gene regulation by 1,25-dihydroxyvitamin D3 and its receptor. J Steroid Biochem Mol Biol 103: 389-395. doi:10.1016/j.jsbmb.2006.12.050. PubMed: 17223545.
- 30. Haussler MR, Haussler CA, Bartik L, Whitfield GK, Hsieh JC et al. (2008) Vitamin D receptor: molecular signaling and actions of nutritional ligands in disease prevention. Nutr Rev 66: S98-112. doi:10.1111/j.1753-4887.2008.00093.x. PubMed: 18844852.
- 31. Ramagopalan SV, Heger A, Berlanga AJ, Maugeri NJ, Lincoln MR et al. (2010) A ChIP-seq defined genome-wide map of vitamin D receptor binding: associations with disease and evolution. Genome Res 20: 1352-1360. doi:10.1101/gr.107920.110. PubMed: 20736230.
- 32. Heikkinen S, Väisänen S, Pehkonen P, Seuter S, Benes V et al. (2011) Nuclear hormone 1α,25-dihydroxyvitamin D3 elicits a genome-wide shift in the locations of VDR chromatin occupancy. Nucleic Acids Res 39: 9181-9193. doi:10.1093/nar/gkr654. PubMed: 21846776.
- 33. Meyer MB, Goetsch PD, Pike JW (2012) VDR/RXR and TCF4/β-catenin cistromes in colonic cells of colorectal tumor origin: iImpact on c-FOS and c-MYC gene expression. Mol Endocrinol 26: 37-51. doi:10.1210/me.2011-1109. PubMed: 22108803.
- 34. Ding N, Yu RT, Subramaniam N, Sherman MH, Wilson C et al. (2013) A vitamin D receptor/SMAD genomic circuit gates hepatic fibrotic response. Cell 153: 601-613. doi:10.1016/j.cell.2013.03.028. PubMed: 23622244.
- 35. Carlberg C, Seuter S, Heikkinen S (2012) The first genome-wide view of vitamin D receptor locations and their mechanistic implications. Anticancer Res 32: 271-282. PubMed: 22213316.
- 36. Carlberg C, Campbell MJ (2013) Vitamin D receptor signaling mechanisms: Integrated actions of a well-defined transcription factor. Steroids 78: 127-136. doi:10.1016/j.steroids.2012.10.019. PubMed: 23178257.
- 37. Gynther P, Toropainen S, Matilainen JM, Seuter S, Carlberg C et al. (2011) Mechanism of 1α,25-dihydroxyvitamin D3-dependent repression of interleukin-12B. Biochim Biophys Acta 1813: 810-818. doi:10.1016/j.bbamcr.2011.01.037. PubMed: 21310195.
- 38. Matilainen JM, Husso T, Toropainen S, Seuter S, Turunen MP et al. (2010) Primary effect of 1α,25(OH)2D3 on IL-10 expression in monocytes is short-term down-regulation. Biochim Biophys Acta 1803: 1276-1286. doi:10.1016/j.bbamcr.2010.07.009. PubMed: 20691220.
- 39. Tsuchiya S, Yamabe M, Yamaguchi Y, Kobayashi Y, Konno T et al. (1980) Establishment and characterization of a human acute monocytic leukemia cell line (THP-1). Int J Cancer 26: 171-176. doi:10.1002/ijc.2910260208. PubMed: 6970727.
- 40. Seuter S, Ryynänen J, Carlberg C (2013) The ASAP2 gene is a primary target of 1,25-dihydroxyvitamin D in human monocytes and macrophages. J Steroid Biochem Mol Biol (in press).
- 41. Seuter S, Heikkinen S, Carlberg C (2013) Chromatin acetylation at transcription start sites and vitamin D receptor binding regions relates to effects of 1α,25-dihydroxyvitamin D3 and histone deacetylase inhibitors on gene expression. Nucleic Acids Res 41: 110-124. doi:10.1093/nar/gkt229. PubMed: 23093607.
- 42. ENCODE Project Consortium, Bernstein BE, Birney E, Dunham I, Green ED, Gunter C, Snyder MENCODE-Project-Consortium, Bernstein BE, Birney E, Dunham I, Green ED, et al (2012) An integrated encyclopedia of DNA elements in the human genome. Nature 489: 57-74. doi:10.1038/nature11247. PubMed: 22955616.
- 43. Thorvaldsdóttir H, Robinson JT, Mesirov JP (2013) Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform 14: 178-192. doi:10.1093/bib/bbs017. PubMed: 22517427.
- 44. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH et al. (2002) The human genome browser at UCSC. Genome Res 12: 996-1006. doi:10.1101/gr.229102. Article published online before print in May 2002. PubMed: 12045153.
- 45. Lee BK, Iyer VR (2012) Genome-wide studies of CCCTC-binding factor (CTCF) and cohesin provide insight into chromatin structure and regulation. J Biol Chem 287: 30906-30913. doi:10.1074/jbc.R111.324962. PubMed: 22952237.
- 46. Schmidt D, Schwalie PC, Wilson MD, Ballester B, Gonçalves A et al. (2012) Waves of retrotransposon expansion remodel genome organization and CTCF binding in multiple mammalian lineages. Cell 148: 335–348. doi:10.1016/j.cell.2011.11.058. PubMed: 22244452.
- 47. Handoko L, Xu H, Li G, Ngan CY, Chew E et al. (2011) CTCF-mediated functional chromatin interactome in pluripotent cells. Nat Genet 43: 630-638. doi:10.1038/ng.857. PubMed: 21685913.
- 48. Giresi PG, Kim J, McDaniell RM, Iyer VR, Lieb JD (2007) FAIRE (Formaldehyde-Assisted Isolation of Regulatory Elements) isolates active regulatory elements from human chromatin. Genome Res 17: 877-885. doi:10.1101/gr.5533506. PubMed: 17179217.
- 49. Song L, Zhang Z, Grasfeder LL, Boyle AP, Giresi PG et al. (2011) Open chromatin defined by DNaseI and FAIRE identifies regulatory elements that shape cell-type identity. Genome Res 21: 1757-1767. doi:10.1101/gr.121541.111. PubMed: 21750106.
- 50. Seuter S, Pehkonen P, Heikkinen S, Carlberg C (2013) The gene for the transcription factor BHLHE40/DEC1/stra13 is a dynamically regulated primary target of the vitamin D receptor. J Steroid Biochem Mol Biol 136: 62-67. doi:10.1016/j.jsbmb.2012.11.011. PubMed: 23220548.
- 51. Seuter S, Pehkonen P, Heikkinen S, Carlberg C (2013) Dynamics of 1α,25-dihydroxyvitamin D3-dependent chromatin accessibility of early vitamin D receptor target genes. submitted..
- 52. Tsuchiya S, Kobayashi Y, Goto Y, Okumura S, Nakae S et al. (1982) Induction of maturation in culture human monocytic leukemia cells by phorbol diester. Cancer Res 42: 1530–1536. PubMed: 6949641.
- 53. Di Rosa M, Malaguarnera G, De Gregorio C, Palumbo M, Nunnari G et al. (2012) Immuno-modulatory effects of vitamin D3 in human monocyte and macrophages. Cell Immunol 280: 36-43. doi:10.1016/j.cellimm.2012.10.009. PubMed: 23261827.
- 54. Ikeuchi T, Nakamura T, Fukumoto S, Takada H (2013) A vitamin D3 analog augmented interleukin-8 production by human monocytic cells in response to various microbe-related synthetic ligands, especially NOD2 agonistic muramyldipeptide. Int Immunopharmacol 15: 15-22. doi:10.1016/j.intimp.2012.10.027. PubMed: 23159604.
- 55. Zhang Y, Leung DY, Richers BN, Liu Y, Remigio LK et al. (2012) Vitamin D inhibits monocyte/macrophage proinflammatory cytokine production by targeting MAPK phosphatase-1. J Immunol 188: 2127-2135. doi:10.4049/jimmunol.1102412. PubMed: 22301548.
- 56. Vanoirbeek E, Krishnan A, Eelen G, Verlinden L, Bouillon R et al. (2011) The anti-cancer and anti-inflammatory actions of 1,25(OH)2D3. Best Pract Res Clin Endocrinol Metab 25: 593-604.
- 57. Adorini L, Penna G (2008) Control of autoimmune diseases by the vitamin D endocrine system. Nat Clin Pract Rheumatol 4: 404-412. doi:10.1038/ncprheum0855. PubMed: 18594491.
- 58. Maxwell PJ, Coulter J, Walker SM, McKechnie M, Neisen J et al. (2013) Potentiation of Inflammatory CXCL8 Signalling Sustains Cell Survival in PTEN-deficient Prostate Carcinoma. Eur Urol 64: 177-188. doi:10.1016/j.eururo.2012.08.032. PubMed: 22939387.
- 59. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N et al. (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3: R34. PubMed: 12184808.