Raji lymphoblastic cell line was kindly provided by Dr. S. E. Dmitriev (Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University). Daudi lymphoblastic cell line was obtained from Official Collection of Somatic Cell lines of D.I. Ivanovsky Institute of Virology of The Ministry of Health and Social Development of The Russian Federation. MCF-7 breast cancer cell line was kindly provided by Dr. E. Zabarovsky, Karolinska Institute (Stockholm, Sweden). MCF-7-2si cell line with suppressed p53 was described previously (17). Raji and Daudi cells were cultured in RPMI 1640 medium (Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum. MCF-7 cells were cultured in DMEM (Life Technologies, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum and 0.01% human insulin. For activation, 100 ng/ml lipopolysaccharide (LPS), 5 ng/ml PMA, and 20 ng/ml ionomycin were added to cell culture for 24 h.

Scientific Council of the Engelhardt Institute of Molecular Biology declared no ethical approval requirements for experiments performed in this study, because only commercially available cell lines were used.

We used PCR to amplify a 155-bp fragment of the CXCR5 promoter (−112/ + 43) containing rs630923. Using two-step PCR, we generated the variants of this fragment containing “A” allele of rs630923 and point mutation of MEF2C-binding site. An additional specificity control represented a DNA fragment from an upstream genome location, which did not contain any MEF2C-binding sites (Table 1). All PCR products were verified by sequencing; sequences of all PCR primers and other synthetic oligonucleotides are represented in Table 1. Nuclear extracts from Raji and Daudi cells were isolated according to the published protocol (18). Nuclear extracts were incubated with DNA fragments and anti-MEF2C antibodies (ab79436, Abcam, Cambridge, UK) and precipitated with pre-blocked protein A sepharose beads. We quantified target DNA by real-time PCR after protein–DNA complexes elution and DNA purification. Background concentrations of the target DNA obtained for control reactions without lysate and without antibodies were subtracted from the data, which was subsequently normalized to the amplicon with mutant MEF2C-binding site. Rabbit IgG antibodies of irrelevant specificity representing an isotype control to the MEF2C antibody produced similar results for all DNA fragments regardless of the presence of the MEF2C-binding site.

Total RNA was extracted from cells using Trizol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. cDNA was generated using M-MULV reverse transcriptase and oligo-dT primer from First strand cDNA synthesis kit (Thermo Scientific, Waltham, MA, USA). Quantitative RT-PCR was performed using Applied Biosystems 7500 real-time PCR amplifier, Real-time PCR reaction mix containing Rox and SYBR Green (Evrogen, Moscow, Russia) and specific primers designed to amplify intron-spanning fragments of human ACTB, CXCR5, and MEF2C genes (Table 1). The PCR program included preheating stage at 95°C for 10 min and 40 cycles of amplification at 95°C for 15 s, 62°C for 20 s, and 72°C for 20 s. Melting curve analysis was used to assess the specificity of amplification. ACTB mRNA coding for human β-actin was used for normalization.

Commercially synthesized single-stranded RNA (Syntol, Moscow, Russia) were annealed by mixing sense and antisense RNA and slow cooling down from 90°C to room temperature in annealing buffer (5×: 50 mM Tris, 100 mM NaCl, pH 8.0). Raji cells were electroporated with siRNA duplexes (500 pmol per 5 million cells) 24 h prior to electroporation with luciferase constructs at which time 200 pmol more of the same siRNA duplex was added to prolong the silencing effect. We used two previously published (19) pairs of siRNAs targeting MEF2C (see Table 1). The second siRNA pair that turned out to be more efficient was used in all subsequent experiments.

Total cell lysates were prepared using 5× Laemmli buffer. Protein samples (2 million cells per sample) were resolved on 12% SDS-PAGE, transferred to Hybond-C Extra nitrocellulose membrane (Amersham Biosciences, Amersham, UK), stained with anti-MEF2C antibodies (ab79436, Abcam, Cambridge, UK) at 1:2000 dilution and with HRP-conjugated secondary anti-rabbit antibody at 1:30,000 dilution. The bands were visualized with ECL using SuperSignal West Dura Extended Duration Substrate (Thermo Scientific, Waltham, MA, USA). Ponceau staining served as a loading control. Gel images were analyzed using ImageJ software to determine the integral intensities of the ECL bands and of the lanes of Ponceau-stained blots.

We amplified the (−455/ + 368) region of human cxcr5 promoter (17, 20) by PCR using genomic DNA from Raji cells as a template and specific primers containing cloning sites. CXCR5 promoter variants containing minor “A” variant of rs630923 or mutation in predicted MEF2C-binding site were generated by two-step PCR mutagenesis and verified by sequencing (see Table 1 for sequences of all oligonucleotides). All variants of CXCR5 promoter were cloned into pGL3-basic luciferase reporter construct (Promega, Madison, WI, USA) using HindIII and NcoI restriction sites.

Cells were electroporated with 5 μg of purified plasmid DNA, and 0.1 μg of pRL-CMV renilla luciferase control reporter vector (Promega, Madison, WI, USA) using Neon Transfection System (Thermo Scientific, Waltham, MA, USA). For each cell line, we used protocols provided by the manufacturer: one 30 ms 1300 V pulse for Raji, two 30 ms 1100 V pulses for MCF-7, and two 30 ms 1300 V pulses for Daudi. Luciferase activity was measured in Luminometer 20/20ⁿ (TurnerBioSystems, Sunnyvale, CA, USA) using Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA) following the manufacturer’s protocol. The activity of Firefly luciferase was normalized to that of renilla luciferase to account for variations in transfection efficiency.

We used Microsoft Excel and Statistical software for statistical analyses. Statistical significance was determined using two-tailed unpaired Student’s t-test, P-value <0.05 was considered significant. Data for each sample represent the result of at least three independent experiments. Real-time PCR and luciferase assays were additionally performed in two technical replicates. Error bars in the graphs represent SEMs value.

Single-nucleotide polymorphism rs630923 (A/C) is located in a highly conservative area of chemokine receptor CXCR5 gene promoter [−31 with respect to the transcription start site (TSS)] and its common “C” allele was shown to be associated with increased risk of MS (12, 15). Since nucleotide variations this close to the TSS may affect the efficiency of transcriptional initiation, we hypothesized that rs630923 may influence CXCR5 promoter activity directly and searched the CXCR5 promoter region for transcription factor binding sites (TFBS) that either appear or disappear depending on the rs630923 allele. We used PERFECTOS-APE software (21) with curated TFBS models from JASPAR and HOCOMOCO databases. The list of candidate transcription factors with best scores included MEF2C, MEF2A, FoxL1, DUX4, HNF4G, and TCF7L2. Of these, MEF2C stood out as the most relevant candidate since it is highly expressed in B-cells (22), participates in inflammation-induced B-cell proliferation (23), and targets a subset of B-cell-specific genes (24). MEF2C was also shown to influence migration potential of leukemic cells by targeting CCR2 and CCR5 chemokine receptors genes (25). Since B-cells facilitate the progression of MS (26), we predicted that the minor “A” rs630923 allele may play its protective role by creating a binding site for MEF2C that attenuates the transcription of CXCR5 gene encoding the critical chemotactic receptor on these cells.

We used pull-down assay to investigate the influence of rs630923 on MEF2C binding to the CXCR5 promoter. We constructed three variants of CXCR5 gene promoter (−112/ + 43) region (Figure 1A) containing either common “C” rs630923 allele, minor “A” allele or a combination of the minor “A” allele with additional mutations disrupting the MEF2C consensus (Figure 1B). Pull-down assay was performed using nuclear extracts from two human lymphoblastic B cell lines, Raji and Daudi (both cell lines carry C/C rs630923 alleles). High levels of CXCR5 and MEF2C expression are typical for both cell lines, according to Broad-Novartis Cancer Cell Line Encyclopedia (27).

A significant amount of the DNA probe containing minor “A” rs630923 allele could be precipitated from the binding reaction with nuclear extracts from both cell lines using anti-MEF2C antibodies while the “C” probe did not significantly differ from the control probes, including the one with point mutation of the MEF2C site (Figure 2). These data demonstrate the presence of a MEF2C-binding site in the CXCR5 promoter containing minor rs630923 allele.

To study the role of rs630923 and associated MEF2C-binding site in modulation of CXCR5 promoter activity, we generated luciferase reporter constructs containing CXCR5 promoter variants both rs630923 alleles with or without the MEF2C-binding site mutation (Figure 3A). In order to emulate BCR stimulation and costimulatory signals received by B cells migrating to MS lesions (28), we stimulated Raji and Daudi B-lymphoblastic cells using combination of LPS, PMA, and ionomycin. B-cell activation leads to increased MEF2C expression (22) but does not influence CXCR5 expression level (29, 30). These previously published observations were verified for CXCR5 by real-time RT-PCR using mRNA from Raji cells (Figure 3B) and for MEF2C by RT-PCR (Figure 3B) and by western blot (Figure 3C).

The presence of either rs630923 “A” allele and/or MEF2C-binding site did not significantly influence the CXCR5 promoter under normal culture conditions, with the exception of the activity of the rs630923 “A” allele variant in Raji cells that was somewhat elevated in this particular set of experiments. However, B-cell activation led to a significant and specific decrease in CXCR5 promoter activity in both B-cell lines when the promoter contained the minor “A” rs630923 variant in the context of a functional MEF2C-binding site (Figure 3A). These data indicate that protective role of “A” rs630923 allele in MS (presumably associated with low CXCR5 and reduced migration of activated B-cells to areas of demyelinating lesions) may be explained by repression of CXCR5 promoter activity by MEF2C transcription factor.

To assess the role of MEF2C transcription factor directly, we evaluated the activities of various CXCR5 promoter variants in Raji cells upon siRNA-mediated suppression of MEF2C expression. We tested two variants of previously published siRNAs against MEF2C (see Materials and Methods), and one of them that caused up to 80% decrease in MEF2C mRNA level and insensitivity of MEF2C expression to cell activation (Figure 4A) was used in subsequent experiments. The activity of CXCR5 promoter lacking MEF2C-binding site (either “C” allele or “A” allele with MEF2C-binding site mutation) was essentially insensitive to both cell activation and MEF2C expression level. However, MEF2C knockdown completely abolished the decrease in the activity of CXCR5 promoter containing the protective “A” allele that was observed in activated cells (Figure 4B). These data indicate that in our cell culture model, both MEF2C and rs630923 protective alleles play unique and essential roles in the reduction of CXCR5 promoter activity during B-cell activation.

It has been suggested earlier that an NFkB binding site in the vicinity of rs630923 may play a role in cxcr5 promoter regulation (15), whereas our own studies of CXCR5 promoter did not reveal any NFkB-binding sites overlapping this SNP (17). To directly access the role of rs630923 in NFkB-dependent CXCR5 promoter activity we used MCF-7 breast cancer cell line and its derivative cell line MCF-7-2si with p53 knockdown (17). This cell culture model takes advantage of the functional antagonism between p53 and NFκB (31) that results in elevated NFκB activity in the absence of p53-mediated suppression (17). Both cell lines showed significant MEF2C expression (Figure 5A). This observation correlated with publically available expression data (27) and indicated that MEF2C effects on CXCR5 promoter could be addressed in this system. The level of NFkB activity, as assessed by an NFkB-dependent reporter plasmid, was significantly higher in MCF-7-2si cells (Figure 5B), again in agreement with our previous results (17). We then measured the activities of the CXCR5 promoter versions containing “C” and “A” alleles of the rs630923 SNP using previously described promoter variant with mutations in all NFkB-binding sites as a negative control (Figure 5C). Both test constructs were significantly more active in MCF-7-2si cells as compared to the parental MCF-7 cell line, but no significant difference between two allelic variants was observed. This result indicates that although CXCR5 promoter activity is modulated by NFkB, rs630923 is not involved in this process.