Caffeine treatment inhibits MICA induction on activated T cells in an NF-κB dependent-manner. (A) PBMCs were pretreated with different doses of caffeine and then stimulated with PHA for 18 hours. Cells were stained with mAbs specific for CD3 and MICA and its expression was evaluated on CD3⁺ T cells. Data are represented as percentage of MICA-positive cells. A representative experiment of 4 is shown. (B) PBMCs were pretreated with caffeine (5 mM) or rapamycin (1 μM) and then stimulated with PHA for 18 hours. Cells were stained with mAbs specific for CD3, MICA (thick line), or control Ig isotype (dashed line). Expression of MICA was evaluated on CD3⁺ T cells. (C) PBMCs were prepared as described in panel B. Data are represented as the mean plus or minus one SD of the percentage of CD3⁺MICA⁺cells of 8 different healthy donors. Significant differences, as calculated by paired t test, are indicated: ***P < .001; ns, not significant. (D) PBMCs were pretreated with 5 mM caffeine and then stimulated with PHA. After 18 hours, CD3⁺ T cells were purified by positive immunomagnetic selection. Total RNA was isolated and used for reverse-transcription PCR reactions with primers specific for MICA and GAPDH. A representative experiment out of 5 is shown. (E) Electrophoretic mobility-shift assay was performed using the ³²P-labeled NF-κB MICA, and the canonical NF-κB Ig oligonucleotide as a probe in the presence of nuclear extracts (10 μg), from unstimulated (−) or PHA-activated PBMCs (3 hours). Where indicated, PBMCs were pretreated with 5 mM caffeine and then stimulated with PHA for 3 hours. The same nuclear extracts were also used with a Octamer factor(s)-specific probe as a control. A representative experiment out of 3 is shown.