The effect of PU.1 knockdown on gene expression and function of mast cells

PU.1 is a hematopoietic cell-specific transcription factor. In the current study, we investigated the role of PU.1 in the gene expression and the function of mouse mast cells (MCs) in vitro and in vivo. When PU.1 siRNA was introduced into bone marrow-derived MCs (BMMCs), IgE-mediated activation was reduced, and the Syk and FcεRIβ mRNA levels were significantly decreased. As the regulatory mechanism of the Syk gene is largely unknown, we performed promoter analysis and found that PU.1 transactivated the Syk promoter through direct binding to a cis-element in the 5′-untranslated region. The involvement of PU.1 in the Syk promoter was also observed in mouse dendritic cells and human MCs, suggesting that the relationship between PU.1 and Syk is common in mammals and in hematopoietic lineages. When antigen was administrated intravenously after the transfusion of siRNA-transfected BMMCs in the mouse footpad, the footpad thickening was significantly suppressed by PU.1 knockdown. Finally, administration of the immunomodulator pomalidomide suppressed passive systemic anaphylaxis of mice. Taken together, these results indicate that PU.1 knockdown might be an efficacious strategy for the prevention of MC-mediated allergic diseases.

FcεRI expression and IgE-mediated degranulation of human MCs 15 . In contrast, it was unclear whether PU.1 knockdown affected the transcription of FcεRI components and the subsequent cell surface expression level and function of FcεRI in mouse MCs. Thus, we evaluated the effect of PU.1 siRNA on the cell surface expression level of FcεRI and the mRNA levels of the FcεRI α−, β−, and γ-chains. First, we evaluated the effect of three siRNAs (#1, #2, and #3) encoding different nucleotide sequences of PU.1. As shown in Fig. 1(a), we confirmed that the three siRNAs significantly knocked down PU.1 mRNA, and siRNA #1 was the most effective. Therefore, we used #1 in the following experiments. Flow cytometric analysis revealed that PU.1 knockdown significantly suppressed cell surface expression of FcεRI on BMMCs when the PU.1 mRNA level decreased under 10% compared with that of the control ( Fig. 1(b)). Although the suppressive effect of PU.1 siRNA on the cell surface FcεRI level was commonly observed in humans 15 and mice ( Fig. 1(b)), surprisingly, PU.1 knockdown decreased the FcεRI β-chain mRNA level, whereas the mRNA levels of the FcεRI αand γ-chains increased in PU.1 knockdown cells ( Fig. 1(c)). Considering that PU.1 knockdown in human MCs decreased the mRNA level of the human α-chain, but did not affect the mRNA levels of human βand γ-chains 15 , the role of PU.1 in the transcription of FcεRI subunits appears to be different between humans and mice. To evaluate the effect of PU.1 knockdown on the expression of signal transduction molecules and on IgE-mediated activation in MCs, we determined the mRNA levels of signal transduction molecules, the degree of IgE-mediated degranulation, and IgE-mediated TNF-α release in BMMCs. Using DNA microarray analysis, we found that Syk mRNA showed the greatest decrease in PU.1 knockdown cells (data not shown). Further detailed analysis using quantitative RT-PCR confirmed that the Syk mRNA level was substantially reduced in PU.1 knockdown cells ( Fig. 1(d)). We also found that transcripts of the phosphatases SHIP-1 and SHIP-2 were markedly increased by PU.1 siRNA knockdown, whereas the mRNA levels of Lyn, PLCγ1, PLCγ2, Fyn, and Stat5 were not affected by PU.1 knockdown ( Fig. 1(d)). Using Western blotting analyses, we confirmed that the protein levels of PU.1, Syk, and FcεRIβ were significantly decreased by PU.1 knockdown ( Fig. 1(e)). The staining of permeabilized cells showed that FcεRIα protein levels in PU.1 knockdown cells were lower than those in control cells ( Fig. 1(f)), suggesting that the mRNA increase in FcεRIα by PU.1 knockdown was not reflected in the total amount of FcεRIα protein ( Fig. 1(c)). The reduction of FcεRIβ protein level may result in suppression of cell surface expression of FcεRIα ( Fig. 1(b)) by affecting the formation and/or stability of the FcεRI complex. After down-regulation of Syk expression and cell surface FcεRI and up-regulation of SHIP-1 and SHIP-2, PU.1 knockdown suppressed IgE-mediated degranulation ( Fig. 1(g)) and TNF-α production ( Fig. 1(h)).

PU.1 directly bound to and transactivated the Syk promoter.
The abovementioned results suggest that PU.1 regulates the Syk expression in MCs. Although Syk plays a key role in signal transduction just downstream of the cell surface receptor, the regulatory mechanism of cell type-specific expression of Syk is largely unknown. Thus, we analyzed the role of PU.1 in Syk expression as follows. First, we excluded the possibility of an off-target effect by the three siRNAs. As shown in Fig. 2(a), the level of Syk mRNA in transfectants was decreased in parallel with the degree of PU.1 mRNA reduction (see Fig. 1(a)). This result demonstrates that suppression of the Syk transcription was due to PU.1 knockdown and not an off-target effect.
PU.1 transactivates target genes in two ways: one is as a monomeric transcription factor, and the other is as a heterodimeric complex with the transcription factor IRF4 or IRF8. To investigate the role of IRF4 and IRF8 in Syk expression, we analyzed the effect of knockdown of IRF4 or IRF8 on Syk expression. Introduction of siRNAs for IRF4 and IRF8 significantly decreased the mRNA levels of IRF4 and IRF8, respectively. In this experimental condition, Syk mRNA and protein levels were maintained ( Fig. 2(b) and (c)), indicating that IRF4 and IRF8 are not involved in Syk expression and that PU.1 activates the Syk promoter as a monomeric transcription factor. Then, we performed a chromatin immunoprecipitation (ChIP) assay to investigate whether PU.1 directly binds to the Syk promoter in MCs. As shown in Fig. 3(a), a significant amount of the chromosomal DNA containing the Syk minimum promoter was immunoprecipitated with anti-PU.1 Ab compared with the control Ab, whereas PU.1 binding was not detected further upstream. Next, luciferase assays using Syk promoter regions of various lengths were performed. As shown in Fig. 3b, the deletion between −125 and −93 markedly reduced luciferase activity in the MC line PT18. A nucleotide replacement at an Ets-motif (−100/−97) significantly decreased the luciferase activity of PT18 transfectants ( Fig. 3(b)). Further luciferase assays with co-expression plasmids showed that exogenous expression of PU.1 transactivated the luciferase activity driven by the wild-type promoter but not that from the mutant promoter lacking the Ets-motif (Fig. 3(c)), suggesting that PU.1 transactivated the Syk promoter through the identified Ets-motif. Furthermore, we performed electrophoretic mobility shift assays (EMSAs) to confirm whether PU.1 directly binds to the Ets-motif at −100/−97 ( Fig. 3(d)). A band shift appeared when the PU.1 protein was added to the reaction mixture containing probe DNA (lane 2). This band disappeared in the presence of excess amounts of a non-labeled wild-type competitor (lane 3, and 4), whereas the specific band remained when a mutant competitor lacking the Ets-motif was used instead of the wild-type competitor (lane 5, and 6), indicating that PU.1 bound to the Syk promoter via the Ets-motif. As the addition of anti-PU.1 Ab resulted in the disappearance of the specific band shift and the appearance of a new band showing lower mobility (lane 8), we confirmed that PU.1 was contained in this complex.
From these results, we concluded that PU.1 transactivates the Syk promoter by directly binding to the Ets-motif at −100/−97, which was identified as a critical cis-element.  BMDCs and found that the Syk mRNA levels were significantly reduced in PU.1 knockdown DCs ( Fig. 4(a)). Furthermore, a ChIP assay showed that a significant amount of PU.1 binds to the above-identified region of the Syk promoter in DCs ( Fig. 4(b)). Although IRF4 and IRF8 are also expressed in DCs, Syk expression was not affected by infection of IRF4 or IRF8 (data not shown), demonstrating that monomeric PU.1 transactivates the Syk gene in DCs in the same manner as that in MCs.

Involvement of PU.1 in the expression of Syk is commonly observed in dendritic cells (DCs) and human cells. Syk is expressed in other immune-related cells including B cells and monocytes
When we compared the nucleotide sequences of the mouse and human Syk genes, we found that the homology of the promoter sequences is not high and that the cis-enhancing element identified in the mouse gene was  Fig. 4(c), these siRNAs effectively knocked down human PU.1 mRNA levels and subsequently reduced Syk mRNA levels in LAD2 cells. In addition, a ChIP assay indicated that PU.1 bound to the proximal region of the human SYK gene in LAD2 (Fig. 4(d)). BMMCs, in which PU.1 siRNA or its control was introduced, into the mouse footpad and determined the footpad thickness before and after an i.v. injection of antigen. The thickness of the control footpad after antigen injection was significantly greater than that before injection, whereas no thickening was observed in the footpad transfused with PU.1 knockdown cells (Fig. 5(a)). These results suggested that PU.1 knockdown is effective for suppression of MC-mediated responses in vivo.
Several studies have reported that the immunomodulatory drug pomalidomide down-regulates PU.1, which is one mechanism by which pomalidomide can be used to treat multiple myeloma [17][18][19] . To evaluate the effect of pomalidomide on expression and function of PU.1 in MCs, we treated LAD2 cells with pomalidomide. Western blotting analysis showed that the protein levels of PU.1 and Syk were decreased in LAD2 cells exposed to pomalidomide for 2-3 days (Fig. 5(b)). These results prompted us to investigate whether pomalidomide has a protective effect on the MC-mediated allergic response in vivo. Then, we analyzed the degree of passive systemic anaphylaxis, which is a well-established mouse model of the IgE-mediated reaction. As shown in Fig. 5(c), decreased body temperature due to the anaphylactic reaction was ameliorated by oral administration of pomalidomide.
These results suggest that down-regulation of PU.1 suppresses the MC-mediated in vivo allergic response.

Discussion
Previously, we found that the transcription factor PU.1 transactivates the FCER1A gene (encoding FcεRIα) and that knockdown of PU.1 reduced expression of FcεRI and subsequently suppressed FcεRI-mediated activation of human MCs 15  Syk is a non-receptor tyrosine kinase that plays an important role in signal transduction initiated by cell surface immunoreceptors, including BCR, FcεRI, FcγR, and C-type lectins 20 . In the present study, we showed that PU.1 is critical for expression of Syk as a transcactivator that directly targets the Syk gene in MCs and DCs. PU.1 expression is observed in hematopoietic cells, especially monocytes, B cells, and neutrophils and is also detected in T cells and MCs. The cell-type specificity of Syk-expressing cells, such as B cells, MCs, neutrophils, macrophages, DCs, osteoclasts, and immature T cells, is similar to that of PU.1 expression. Although we used MCs and DCs to demonstrate the involvement of PU.1 in Syk expression, the role of PU.1 in Syk expression may be common in other hematopoietic lineages, such as B cells. Syk inhibitors exhibit therapeutic effects on allergic diseases, autoimmune diseases, and B lymphocyte malignancies 20 . Therefore, PU.1 suppression may be useful for preventing these diseases by inhibiting Syk-mediated signaling accompanied by suppressing the expression of several cell type-specific genes.
In addition to Syk reduction, PU.1 knockdown suppressed the expression of cell surface FcεRI due to a decrease in Ms4a2 mRNA (encoding FcεRIβ), along with a substantial increase in SHIP-1 and SHIP-2. In classical promoter analyses, GATA1 and FOG-1 were identified as transcriptional regulators of the mouse Ms4a2 gene 9,16 , whereas Oct-1 and MZF-1 were originally identified as regulators of the human MS4A2 gene 10,13 . This difference is likely because the nucleotide sequences of the promoters are not conserved between humans and mice. Recently, we demonstrated that the mRNA level of the human MS4A2 gene was decreased in GATA2 siRNA-introduced MCs but was not affected by PU.1 siRNA 15 . Therefore, the down-regulation of Ms4a2 mRNA by PU.1 knockdown is a mouse-specific observation. Further detailed analysis is required to clarify the mechanism underlying PU.1 involvement in the transcription of the Ms4a2 gene, for instance, as a direct transactivator or an indirect regulator through transcription of another factor. PU.1 knockdown has not been found to up-regulates SHIP-1 expression in any cells thus far. Interestingly, several transcription factors were identified as suppressive regulators of SHIP gene expression; for example, Ikaros binds to the promoter of the SHIP-1 gene in B cells, and a deficiency of Ikaros up-regulates SHIP-1 expression 21 , whereas Fli-1 suppresses SHIP-1 transcription in erythroleukemia 22 . Although the molecular mechanism underlying how these hematopoietic cell-specific transcription factors function as suppressors of the SHIP-1 gene is largely unknown, detailed analysis regarding of the role of PU.1 in SHIP-1 expression may clarify this issue. Regardless, we concluded that PU.1 knockdown suppressed FcεRI-mediated signaling with a decrease in positive regulators (Syk and FcεRI) and an increase in suppressors (SHIPs).
In the present study, we used two models to demonstrate that PU.1 knockdown suppressed in vivo allergic responses. In the first experiment, to evaluate the effect of MC-specific knockdown of PU.1, we transfused siRNA-pre-injected MCs into mice. Although MC-deficient mice, such as Wsh/sh, would be a better recipient to exclude the involvement of endogenous MCs, we showed a significant effect of PU.1 siRNA on the MC-mediated response in vivo, even in this experimental condition. As a preliminary experiment, we examined the degree of anaphylactic reaction of Wsh/sh mice injected with PU.1 siRNA-introduced MCs or control cells and found that the degree of rapid body temperature decrease was reduced by MC-specific knockdown of PU.1 (data not shown). For the second approach, we administered the immunomodulatory drug pomalidomide to mice. Although pomalidomide reduced the protein levels of PU.1 and Syk in human MC cells, several issues remain to be clarified, such as the possibility that pomalidomide modulates the expression of other molecules involved in the activation of MCs. In addition, whether the protein levels of PU.1 and Syk in MCs were reduced by the pomalidomide treatment should be examined. Further detailed analysis regarding the effect of pomalidomide on allergic responses is required.
We demonstrated that PU.1 knockdown significantly reduced the MC-mediated allergic response in vivo and in vitro. The development of nuclear medicine, which can specifically and effectively deliver siRNA or antisense oligonucleotide to target cells, is required to further evaluate the efficacy of PU.1 knockdown for the treatment of immune-related diseases. We will investigate drug delivery systems for nuclear medicine to specific immune cells in future studies.

Materials and Methods
Mice and cells. BMMCs were generated from bone marrow cells of C57BL/6 mice (Japan SLC, Hamamatsu, Japan) by maintenance in RPMI 1640 (Sigma-Aldrich, St. Louis, MO) supplemented with 10% heat-inactivated fetal calf serum, 100 U/mL penicillin, 100 μg/mL streptomycin, 100 μM 2-mercaptoethanol, 10 μM minimum essential medium nonessential amino acid solution, and 5 ng/mL of murine IL-3 (PeproTech, London, United Kingdom) at 37 °C for more than 5 weeks. BMDCs were obtained by 10-day culture of BM cells in the medium The plasmid pCR3-mPU.1 or mock vector (pCR3.1) was introduced into cells with an internal control plasmid, pRL-CMV (Promega). Transfection of BMMCs and PT18 was performed using a Neon system in the same way as for siRNA transfection, and calcium phosphate was used for transfection of CV-1 and HEK293T cells. Briefly, plasmid DNA in 0.25 M CaCl 2 solution was mixed with an equal volume of 2XHBSS, and the mixture was added dropwise to the cells after incubation for 20 min at room temperature. Determination of luciferase activity was performed as previously described using a 1420 Luminescence Counter ARVO Light (Perkin Elmer) 25 . For determination of β-galactosidase activity, cell lysates were incubated with a substrate, o-nitrophenyl-β-D-galactopyranoside (ONPG) (Thermo Scientific, Waltham, MA), for the appropriate time at 37 °C, and the absorbance of the reaction mixture at 405 nm was measured.
Western blot analysis. Western blot analyses were performed as described previously 24,28 . Antibodies against Syk (#LR), FcεRIβ (#N-18), IRF4 (#M-17), and IRF8 (#C-19) were purchased from Santa Cruz Biotechnology, and anti-β-actin Ab (#AC-15) was from Sigma-Aldrich. Anti-PU.1 Ab was the same as that used in ChIP assays. Passive systemic anaphylaxis. Mice were orally administered 10 mg/kg/day of pomalidomide or saline for 6 days. On day 6, mice were injected intravenously with 3 μg/ml of TNP-specific IgE in 200 μl of saline and then injected with 200 μg of TNP-BSA intravenously at 5 h after IgE injection. The body temperature of each animal was measured every 15 min for 1 h after antigen injection. Statistical analysis. Statistical analysis was performed using a two-tailed Student's t-test with p values < 0.05 considered to be significant.