The human HEK 293 T and Jurkat cells were obtained from the German collection of microorganisms and cell cultures (DSMZ) and authenticated using STR DNA typing. HEK 293 T and Jurkat cells were cultured as described previously [14]. All cell lines were cultured for less than six month after receipt.

CD4⁺ and CD8⁺ T cells were isolated by negative selection, purity was confirmed by flow cytometry (Additional file 1: Figure S2) and cultured in RPMI 1640 medium as mentioned earlier [20].

2.5 × 10⁵ Jurkat cells/2 ml/6well or 1 × 10⁶ CD4⁺ T cells/ml/12well or 1 × 10⁶ CD8⁺ T cells/ml/12well were transfected with 150 ng hsa-miR-34a-5p miScript miRNA mimic (MIMAT0000255: 5’UGGCAGUGUCUUAGCUGGUUGU), or the allstars negative control (ANC) using HiPerFect transfection reagent (Qiagen, Hilden, Germany). 48 h post transfection, cells were harvested and whole cell extracts were prepared as described above and the CD4⁺ and CD8⁺ T cells were stained with anti-CD4-FITC (RPA-T4, BD), with anti-CD8-FITC (RPA-T8, BD) and anti-CD11A-APC (HI111, BD), or respective conjugated isotype control antibodies, fixed in 1% paraformaldehyde and analyzed by flow cytometry (FACS canto II, BD biosciences)

For the dual luciferase reporter assays 7 × 10⁴ HEK 293 T cells per well of a 24-well plate were transfected with 200 ng/well reporter vector and 800 ng/well miR-34a expression plasmid using PolyFect transfection reagent (Qiagen, Hilden, Germany) corresponding to manufacturer’s protocol. Dual Luciferase assays were performed as mentioned earlier and according the manufacturer’s protocol [14]. For analysis the luciferase activity of each wild type 3’UTR reporter construct cotransfected with miR-34a was normalized to the luciferase activity of the empty reporter vector cotransfected with miR-34a.

For Western Blot analysis of CD11A and VAMP2 Jurkat, CD4⁺ T cells, or CD8⁺ T cells were transfected as described above. 48 h post transfection cells were lysed with 2x lysis buffer (130 mM Tris/HCl, 6% SDS, 10% 3-Mercapto-1,2-propandiol, 10% glycerol) and 3 times treated with ultrasound for 2 s. 15 μg of the whole protein extracts were separated by SDS gel electrophoresis in a Mini-Protean® TGX Precast Gel (Bio-Rad Laboratories Inc., Hercules, California, USA) and transferred to a nitrocellulose membrane (Whatman, GE Healthcare, Freiburg, Germany). CD11A was detected by a purified mouse anti human CD11A antibody (610826,BD, Franklin Lakes, USA), VAMP2 by a monoclonal rabbit anti human VAMP2 antibody (D601A, Cell Signaling Technology, Danvers, United States). GAPDH and β-actin were used as loading controls and detected with a monoclonal antibody against human GAPDH (14C10, Cell Signaling Technology, Danvers, United States) and an anti-β-actin monoclonal mouse antibody (AC-15, Sigma Aldrich, Munich, Germany), respectively. All secondary antibodies were obtained from Sigma Aldrich (Sigma Aldrich, Munich, Germany).

The pSG5-miR-34a expression vector was generated by Eurofins Genomics containing the nucleotides 9151617–9151816 of chromosome 1 (Eurofins Genomics, Ebersberg, Germany). The 3’UTRs of VAMP2, IKBKE, MYH9, MARCH8, KLRK1, CD11A, TRAFD1, CCR1, PYDC1, PRF1, PIK3R2, PIK3CD, AP1B1, ADAM10, PVR, AP2S1, BAD, ICOS, CD247, ZFP36, STX8 and SPN, were cloned into the pMIR-RNL-TK vector, which was described in Beitzinger et al. using the SpeI, SacI or NaeI restriction sites [21]. All insert were PCR amplified using specific primers and all predicted hsa-miR-34a-5p target sites of selected target genes were mutated by site-directed mutagenesis with the QuickChange II Site-Directed Mutagenesis Kit (Agilent Technologies, Santa Clara, United States) using specific primers. The identifiers of all cloned 3’UTR sequences and the sequences of specific cloning primers are shown in Additional file 1: Table S1.

The RNA isolation of ANC or miR-34a transfected CD4⁺ T cells was carried out 48 h post transfection using the miRNeasy Mini Kit corresponding to the manufacturer’s protocol (Qiagen, Hilden, Germany). The expression of hsa-miR-34a-5p, was analyzed applying qRT-PCR with the StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, United States) and the miScript PCR System (Qiagen, Hilden, Germany) according to the manufacturer’s manual. In brief, 200 ng total RNA was reverse transcribed into cDNA using the miScript RT II Kit with the miScript HiFlex Buffer (Qiagen, Hilden, Germany). RNU48 served as endogenous control for miRNA expression. Over-expression of miR-34a in the transfected CD4⁺ T cells is shown in Additional file 1: Figure S1.

1 × 10⁶ CD8⁺ T cells/ml/12well were transfected with 150 ng hsa-miR-34a-5p miScript miRNA Mimic (MIMAT0000255: 5’UGGCAGUGUCUUAGCUGGUUGU), or the allstars negative control (ANC) using HiPerFect transfection reagent (Qiagen, Hilden, Germany). 48 h post transfection the transfected CD8⁺ T cells were activated by PMA/Ionomycin. 4 h after activation the supernatants were collected and PRF1 quantification was performed according to the manual of the human Perforin ELISA Kit (#PK-EL-68242, PromoCell GmbH, Heidelberg, Germany).

Statistical analysis of the luciferase assays, the Western Blots, the FACS analysis and ELISA was performed with SigmaPlot 10 (Systat, Chicago, USA) applying Student’s t-test. Quantification of the Western blots was carried out with Image Lab Software Version 5.2.1 (Bio-Rad Laboratories Inc., Hercules, California, USA).

Previously, we identified miR-34a as modulator of intracellular calcium signaling and NF-κB signaling in CD4⁺/CD8⁺ T cells [19, 20]. To investigate the overall importance of miR-34a in T cell regulation we performed an in silico prediction of target genes of miR-34a using miRWalk 2.0 [22] and identified 18828 potential target genes of miR-34a. miRWalk 2.0 combined 10 algorithms including DIANAmT, miRanda, miRDB, miRWalk, RNAhybrid, PICTAR4, PICTAR5, PITA, RNA22 and Targetscan. By including only genes that were predicted by at least 4 different target prediction algorithms, we reduced the number of target genes to 3179. To arrange the predicted target genes in pathways we used GeneTrail2 (https://genetrail2.bioinf.uni-sb.de/), a web service allowing the integrated analysis of transcriptomic, miRNomic, genomic and proteomic datasets [23]. We identified 1227 significant subcategories (p value ≤0.05) in Gene Ontology. We analyzed all subcategories for immune system related pathways and found the highest number of predicted miR-34a target genes in the subcategory immune system process with 193 predicted miR-34a target genes that were significantly enriched in this pathway (p value ≤0.05) (Additional file 1: Table S2). This list was refined by deleting 29 target genes, which were already validated by others using miRTarBase [24] and 4 target genes, which were previously verified by us [19, 20] (Additional file 1: Table S3). Out of the remaining 160 predicted target genes we selected 22 miR-34a-target genes for experimental analysis based on their predicted biological function according to the Gene Ontology (GO) knowledgebase. Figure 1a depicts the detailed affiliation of the target genes in the specialized subcategories of the immune system process category as indicated in the Gene Ontology database.Fig1Fig. 1

Gene Ontology subcategories of the predicted miR-34a target genes. a Enrichment of predicted miR-34a target genes in specific Gene Ontology subcategories. b Dual luciferase reporter gene assay of VAMP2, IKBKE, MYH9, MARCH8, KLRK1, CD11A, TRAFD1, CCR1, PYDC1, PRF1, PIK3R2, PIK3CD, AP1B1, ADAM10, PVR, AP2S1, BAD, ICOS, CD247, ZFP36, STX8 and SPN. HEK 293 T cells were co-transfected with miR-34a and reporter plasmids containing 3’UTRs of target genes as indicated. The luciferase activities were normalized with respect to the luciferase activity measured with empty reporter construct. The results represent the mean of four independent experiments carried out in duplicates. Columns colored in turquois represent a significant reduction of the luciferase activity with a p-value ≤0.001 (three asterisks). Columns colored in magenta represent a significant reduction of the luciferase activity with a p-value ≤0.01and ≥ 0.001 (two asterisks). Columns colored in violet represent a significant reduction of the luciferase activity with a p-value ≤0.05 and ≥ 0.01 (one asterisk). Columns colored in dark blue represent a non-significant reduction of the luciferase activity with a p-value ≥0.05. Data are represented as mean ± SEM

By our in silico target prediction we identified miR-34a binding sites in the 3′ UTRs of VAMP2, IKBKE, MYH9, MARCH8, KLRK1, CD11A, TRAFD1, CCR1, PYDC1, PRF1, PIK3R2, PIK3CD, AP1B1, ADAM10, PVR, AP2S1, BAD, ICOS, CD247, ZFP36, STX8 and SPN. The sequences, the positions within the 3’UTRs as well as the types of the miR-34a binding sites are shown in Table 1. We amplified the nucleotides of the miR-34a binding sites by PCR and cloned this PCR product into the pMIR-RNL-TK reporter vector. The cloned reporter constructs were used in a dual luciferase reporter assays. To this end, the reporter plasmids or the empty reporter vector were co-transfected with an empty pSG5 plasmid or a miR-34a expression vector in HEK 293 T cells. The luciferase activities of the co-transfections with reporter constructs harboring the predicted 3’UTRs and miR-34a expression plasmid were normalized with regards to the luciferase activities of the co-transfections with empty reporter vector and miR-34a expression plasmid. We found the strongest reduction of the luciferase activity for the VAMP2 reporter plasmid that showed an activity of only 49% (p value≤0.001) when co-transfected with miR-34a. Likewise, the luciferase activities of the reporter construct for IKBKE, MYH9, MARCH8, KLRK1, CD11A, TRAFD1, CCR1, PYDC1, PRF1, PIK3R2, PIK3CD, AP1B1, ADAM10, PVR, AP2S1 and BAD were each significantly decreased (Fig. 1b). In detail, the luciferase activity of IKBKE reporter vector was decreased to 53%, of MYH9- to 54%, of MARCH8- to 62%, of KLRK1- to 67%, of CD11A- to 68%, of TRAFD1- to 70%, of CCR1- to 71%, of PYDC1- to 74%, of PRF1- to 76%, of PIK3R2- to 78%, of AP1B1- to 81%, of ADAM10- to 81%, of PVR- to 82%, of AP2S1- to 90%, and the activity of BAD-reporter vector to 91%. The reporter constructs of ICOS, CD247, ZFP36, STX8 and SPN showed no significant reduction of the luciferase activity. To verify the binding of miR-34a to its target sites we mutated all binding sites in the 3’UTRs of VAMP2, IKBKE, MYH9, MARCH8, KLRK1, CD11A, which displayed a distinct decrease of the luciferase activity as well as all binding sites in the 3’UTRs of ADAM10, and PVR, which showed only a slight reduction. We could validate the direct binding of miR-34a to its binding sites in the 3’UTRs of VAMP2, IKBKE, MYH9, MARCH8, KLRK1, CD11A and ADAM10 showing a significant increase of the luciferase activity of the mutated reporter constructs in comparison to their wild type 3’UTRs (Fig. 2). For PVR we failed to provide evidence that miR-34a directly binds to its predicted binding site. The dual luciferase assays were done in duplicates and have been repeated 4 times.Tab1

Schematic representation of the reporter gene plasmids

We investigated the downstream effect of miR-34a binding to the 3’UTRs of VAMP2 and CD11A on their endogenous protein levels in the Jurkat cell line, in primary CD4⁺ and CD8⁺ T cells. VAMP2 was chosen for further analysis as most affected miR-34a target gene in the dual luciferase assay and CD11A due to its pivotal role in the anti-tumor response and T cell activation. Purity of isolated CD4⁺ and and CD8⁺ T cells were analyzed by flow cytometry (CD4⁺ T cells: mean 91.1% ± 2.5% in three independent experiments, CD8⁺ T cells: mean 91.25% ± 0.9% in three independent experiments). We transfected Jurkat, primary CD4⁺ and CD8⁺ T cells either with “allstars negative control” (ANC) as a non-targeting control or with miR-34a-5p mimic. The over-expression of miR-34a in the transfected CD4⁺ T cells was confirmed by qRT-PCR as shown in Additional file 1: Figure S1. Using specific antibodies against VAMP2 or CD11A we analyzed the endogenous protein levels by Western blotting and found reduced levels of both endogenous VAMP2 and CD11A in the miR-34a transfected Jurkat, CD4⁺ T cells and CD8⁺ T cells (Fig. 3a-f). Representative Western blots out of 3 independent experiments are shown in Fig. 3a-f. Figures 3g-l depict the quantifications of the endogenous VAMP2 and CD11A protein levels for all experiments in Jurkat, CD4⁺ and CD8⁺ T cells. The results show that the mean VAMP2 protein level levels were reduced upon transfection of miR-34a in Jurkat cells to 54% (p value≤0.01) (Fig. 3g), in CD4⁺ T cells to 51% (p value≤0.05) (Fig. 3h) and in CD8⁺ T cells to 56% (p value≤0.001) (Fig. 3i). Mean CD11A protein levels were reduced upon transfection of miR-34a in Jurkat cells to 78% (p value≤0.05) (Fig. 3j) and in CD4⁺ T cells to 75% (p value≤0.05) (Fig. 3 k) and in CD8⁺ T cells to 48% (p value≤0.05) (Fig. 3 l).Fig3Fig. 3

Western blot analysis of VAMP2 and CD11A. a-c Western blot analysis of VAMP2 in miR-34a transfected Jurkat (a), CD4⁺ (b) and CD8⁺ T cells (c). The cells were transfected either with allstars negative control (ANC) or miR-34a-5p mimic. 48 h after transfection the endogenous protein level of VAMP2 was analyzed by Western blotting using specific antibodies against VAMP2. GAPDH served as loading control. d-f: Western blot analysis of CD11A in miR-34a transfected Jurkat (d), CD4⁺ (e) and CD8⁺ T cells (f). The cells were transfected either with allstars negative control (ANC) or miR-34a-5p mimic. 48 h after transfection the endogenous protein level of CD11A was analyzed by Western blotting using specific antibodies against CD11A. Beta actin served as loading control in Jurkat cells and CD4⁺ T cells. GAPDH served as loading control in CD8⁺ T cells. g-i: Quantification of endogenous VAMP2 protein levels in miR-34a transfected Jurkat (g), CD4⁺ (h) and CD8⁺ T cells (i). Three independent Western Blots were quantified by densitometry using Image Lab Software. The protein expression of VAMP2 was normalized with respect to the corresponding GAPDH signals of the appropriate samples. One asterisk corresponds to a p-value ≤0.05 and ≥ 0.01, two asterisks to p-value ≤0.01 and ≥ 0.001 and three asterisks to p-value ≤0.001 . j-l: Quantification of endogenous CD11A protein levels in miR-34a transfected Jurkat (j), CD4⁺ (k) and CD8⁺ T cells (l). Three independent Western Blots were quantified by densitometry using Image Lab Software. The protein expression of CD11A was normalized with respect to the corresponding beta actin or GAPDH signals of the appropriate samples. One asterisk corresponds to a p-value ≤0.05 and ≥ 0.01

For functional downstream analysis of miR-34a over-expression in CD8⁺ T cells, we analyzed the PRF1 (Perforin 1) secretion of activated CD8⁺ T cells transfected either with “allstars negative control” (ANC) as a non-targeting control or with miR-34a-5p mimic. Four hours post activation, the secretion of PRF1 of these cells was determined by a PRF1 ELISA. Figure 4 E depicts the quantification of PRF1 in supernatants of control or miR-34a-5p transfected CD8⁺ T cells in three independent experiments from 2 different donors. Mean PRF1 levels decreased upon transfection of miR-34a in CD8⁺ T cells to 49% (p value≤0.001) (Fig. 4e).