Fig 2: PRMT6 is a NF-κB coactivator. (A) HEK 293 cells were transiently cotransfected with a NF-κB firefly luciferase reporter plasmid and a Renilla luciferase construct along with a GFP vector (Mock), a GFP-CARM1 or a GFP-PRMT6 construct. Cells were left unstimulated or stimulated 20 h after transfection with TNF-α (10 ng/ml, 6 h). Error bars represent standard deviation calculated from quadruplicate luciferase assays. Mean values are expressed as fold changes in luciferase activity with the unstimulated Mock arbitrarily set as 1. The P values obtained for PRMT6 and CARM1 transfections compared to the Mock are 0.00016 and 0.0012, respectively. These data are representative of four independent experiments. (B) HEK 293 cells were transiently cotransfected with a NF-κB luciferase reporter plasmid and a Renilla luciferase construct along with an empty vector or pMyc-PRMT6 or a PRMT6 catalytic inactive form (dead). Cells were unstimulated or stimulated 20 h after transfection with TNF-α (10 ng/ml, 6 h). Error bars represent standard deviation calculated from quadruplicate luciferase assays. Mean values are expressed as fold changes in luciferase activity with the unstimulated Mock arbitrarily set as 1. The P values obtained for the TNF-α-treated groups of PRMT6 WT and dead enzymes are 0.0006 and 0.43, respectively, when compared to the Mock. (C) HeLa cells were transfected with GFP (Mock), GFP-PRMT6 or GFP-CARM1 constructs for 24 h and then stimulated with TNF-α (10 ng/ml, 60 min). Total RNA was analyzed by quantitative RT-PCR for the expression of the indicated genes and normalized against GAPDH. For each gene, the TNF-α-stimulated Mock group mean was arbitrarily set as 1. Error bars represent standard deviation calculated from triplicates. For IL-6, the P values of GFP-PRMT6 and GFP-CARM1 versus Mock are 0.005 and 0.0004, respectively. For TNF-α, the P values of GFP-PRMT6 and GFP-CARM1 versus Mock are 9.2E-07 and 0.002 respectively. For MCP1, the P values of GFP-PRMT6 and GFP-CARM1 versus Mock are 0.3 and 0.07, respectively. For COX-2, the P values of GFP-PRMT6 and GFP-CARM1 versus Mock are 5.8E-05 and 0.64, respectively. (D) WT and PRMT6-KO immortalized Mefs were grown to 90% confluency, at which point cells were unstimulated or stimulated with TNF-α (10 ng/ml, 60 min). Total RNA was isolated and analyzed by quantitative RT-PCR for the expression of the IL-6 gene and normalized against GAPDH. Error bars represent standard deviation calculated from triplicates. The mean value for the unstimulated groups was arbitrarily set as 1. The P values of the TNF-α-stimulated groups compared to the unstimulated are 0.0002 and 0.0024 for the WT and KO cells, respectively. The P value of the fold decrease difference for KO versus WT cells is 0.0014.

Fig 3: RelA coimmunoprecipitates with endogenous PRMT6 and they directly interact in vitro. (A) Primary MEFs were isolated from WT and ER*-PRMT6 Line A mice, cultured for 7 days and treated with OHT (2 μM) for 24 h. Cells were then lysed and an IP was performed using αFlag agarose. A WB using αRelA antibody reveals an interaction between the ER*-PRMT6 and the endogenous RelA in the lysate from the transgenic cells. (B) HEK 293 cells were grown to 90% confluency, then washed with PBS and lysed in co-IP buffer. Lysates were divided in equal volumes and incubated with αIgG or αPRMT6 antibodies overnight and with protein A/G agarose thereafter. A WB using αRelA antibody reveals an interaction between PRMT6 and RelA. (C) Immortalized WT and PRMT6-KO MEFs were grown to 90% confluency, and then washed with PBS and lysed in co-IP buffer. Lysates were incubated with αPRMT6 antibody (which recognized mouse PRMT6) overnight and with protein A/G agarose thereafter. A WB using αRelA antibody reveals an interaction between PRMT6 and RelA in the WT cells. (D) Direct binding between PRMT6 and RelA. GST and GST-RelA (1–431) were incubated with His-PRMT6 and a GST pull-down was performed. For an additional negative control, GST-RelA was also incubated with a His-Tudor-UHRF1. A WB using αHis antibody reveals direct binding between GST-RelA and His-PRMT6.

Fig 4: The inducible ER*-PRMT6 fusion is functional in vivo. (A) Three transgenic lines were established using the pCAGGS-ER*-PRMT6 vector. Southern analysis was performed with a PRMT6 cDNA probe. The arrowhead indicates the transgene. (B) Intraperitoneal injection of 1-mg Tamox, for 5 consecutive days, induced stabilization of ER*-PRMT6 in lung, brain and liver of Line A mice. The vehicle (sunflower oil) was injected as a control. (C) IF for the Flag-tag performed on skin sections from transgenic Line A mice shows ER*-PRMT6 stabilization in the nucleus when Tamox was applied topically on shaved dorsal skin (1 mg/day) for 5 days. (D) Core histones were isolated from epidermal scrapes of topically Tamox-treated mice (1 mg every other day for 10 weeks) and subjected to WB with an αH3R2me2a antibody. (E) The ER*-PRMT6 Line A mice die upon prolonged Tamox administration. A survival curve was generated after topical Tamox administration (1 mg/day) until death of all transgenic mice occurred (n = 14). WT mice were not impacted by this treatment. (F) IL-6 serum levels after 12 topical applications of Tamox (1 mg/day) were measured by ELISA in four groups of animals (WT, WT Tamox-treated, Transgenic (Tg) Line A, Tg Line A Tamox-treated). Wilcox and Student t-tests were run and a P value of 0.004 was obtained for the Tg Tamox-treated group (n = 32) when compared with the WT Tamox-treated group (n = 20) for both tests. All the mice in this study were healthy at the time of serum harvesting. All mice were between 8 and 9 weeks of age, and females and males were equally represented.

Fig 5: Characterization of an inducible ER*-PRMT6 fusion. (A) Human PRMT6 cDNA was cloned into the pCAGGS vector, downstream an ER* (a truncated version of the estrogen receptor that binds Tamox). A Flag-tag was introduced between the two proteins. The ubiquitous β-actin promoter drives the expression of the chimeric ER*-PRMT6 protein. (B) Graphic depiction of this approach. ER*-PRMT6 is localized in the cytoplasm. Upon Tamox or OHT treatment, the chimera protein becomes stabilized and translocates into the nucleus. (C) HEK293 cells stably transfected with pCAGGS-ER*-PRMT6 were treated with OHT (2 μM) and then separated into nuclear [N] and cytoplasmic [C] fractions. Western analysis was performed using an αFlag antibody to detect ER*-PRMT6. An αLamin A/C Western was performed to confirm the quality of the nuclear/cytoplasmic fractionation. Time points after OHT treatment are indicated. (D) Core histones were isolated from the same ER*-PRMT6 HEK 293 cells shown in (C). The core histones were subjected to western analysis with an αH3R2me2a antibody to monitor accumulation of this mark. Equal loading was confirmed by Ponceau staining and αH3 western analysis.