SB203580 Enhances LPS-Induced G-CSF Production in Macrophages by Increasing G-CSF mRNA Stability

顆粒性白血球群落刺激性因子 (Granulocyte colony-stimulating factor, G-CSF)是造血醣蛋白生長因子的成員之一，具有調控嗜中性白血球分化增生的能力。G-CSF可以經過LPS、IL-1β及TNF-α的刺激透過轉錄及後轉錄調控機制由內皮細胞、巨噬細胞及其他許多免疫細胞分泌。G-CSF mRNA三端未轉譯區 (3’ untranslated region, 3’UTR) 含有數個AUUUA序列 (adenosine uridine-rich elements, ARE) 及莖環不穩定元素 (stem loop destabilizing element, SLDE) 與mRNA的不穩定有關。有多個發炎細胞激素的3’UTR可以發現都有ARE，多篇文獻報導在p38 MAPK 訊息傳遞路徑活化下，這些含有ARE的發炎細胞激素mRNA可以由不穩定狀態變成穩定的狀態。p38α被發現與調控轉譯及後轉錄進而影響細胞激素的合成較具關聯性，p38 MAPK抑制劑SB203580可以抑制p38α與β藉由降低mRNA的半生期進而降低發炎細胞激素的產生。然而，SB203580在巨噬細胞中G-CSF的表現影響仍然是未知的。 在我們的研究當中發現，在事先經SB203580處理三十分鐘後，再以LPS刺激的RAW264.7巨噬細胞及小鼠骨髓分化的巨噬細胞(BMDM)都導致G-CSF mRNA及蛋白質的表現增加，但是會抑制LPS所誘導IL-1β mRNA的表現，可以看到G-CSF與IL-1β有不一樣的反應。因此，想了解在SB203580/LPS處理的細胞G-CSF mRNA表現增加的調控機制，利用luciferase reporter assay分析G-CSF啟動子活性以及以nuclear run on assay測mRNA轉錄合成的速率，結果得知LPS可以使G-CSF啟動子活性及轉錄上升，然而加入SB203580並沒有促進G-CSF轉錄的作用。接著，我們探討SB203580是否會影響G-CSF mRNA的穩定度，對已經以LPS或SB203580/LPS處理的細胞以actinomycin D抑制轉錄後，以Northern blot看mRNA的變化，發現SB203580/LPS處理的細胞G-CSF mRNA的穩定性比起只有LPS刺激的來的穩定。為了研究細胞在SB203580/LPS處理之下G-CSF mRNA穩定度高是哪一段序列所媒介的，於是將G-CSF 5’UTR、3’UTR及coding region接至pGL3-control質體的luciferase基因前面或後面，分別為pGL3-G-CSF 5’UTR-Luc、pGL3-Luc-G-CSF 3’UTR、 pGL3-Luc-G-CSF(+35/+372) 及 pGL3-Luc-G-CSF(+340/+720)。將這些質體轉染入RAW264.7後，事先以SB203580處理或不處理，再以LPS刺激後測luciferase活性，發現轉染pGL3-G-CSF 5’UTR-Luc及pGL3-Luc-G-CSF 3’UTR兩種質體luciferase活性很低(與pGL3-control相比)，在SB203580/LPS處理之下luciferase活性並沒有比LPS處理的增強；當細胞轉染pGL3-Luc-G-CSF(+35/+372) 及 pGL3-Luc-G-CSF(+340/+720) luciferase活性與pGL3-control差不多，但在SB203580/LPS處理之下與pGL3-control相比 luciferase活性些微上升。這些結果顯示，G-CSF 5’UTR及3’UTR對mRNA具不穩定的影響，而G-CSF coding region對mRNA則有穩定的功能。最後，利用p38 siRNA致弱p38的表現，以LPS刺激後發現G-CSF的表現與negative control siRNA對照組相比較下是降低的。綜合以上結果，我們推論p38 MAPK抑制劑SB203580會提高G-CSF mRNA及蛋白質的量，可能是因為SB203580提高G-CSF mRNA穩定性的結果，但這些作用似乎不是因為SB203580抑制p38所造成的。The granulocyte colony-stimulating factor (G-CSF) is a member of the glycoprotein growth factor family that controls the proliferation and differentiation of the neutrophilic granulocyte lineage. G-CSF is produced by endothelium, macrophages, and a number of other immune cells. Expression of G-CSF is induced by several inflammatory stimuli such as lipopolysaccharide (LPS), interleukin-1β (IL-1β) and tumor necrosis factor alpha (TNF-α) through transcriptional and post-transcriptional mechanisms. The G-CSF mRNA contains adenosine uridine-rich elements (AREs) and stem loop destabilizing element (SLDE) in the 3’ untranslated region (3’UTR) which have been associated with mRNA instability. AREs are found in the 3’UTRs of many inflammatory cytokines. Numerous reports demonstrated that AREs act as mRNA destabilization determinants but also confer LPS-induced stabilization of the mRNA through p38 MAPK dependent pathway. The p38α has been found to regulate cytokine biosynthesis in both translation and mRNA stability. Pyridinyl imidazole compounds, such as SB203580, inhibit p38α and β and block production of several major inflammatory cytokines, such as TNF-α, IL-1β, IL-6 and granulocyte macrophage-colony stimulating factor (GM-CSF), by decreasing their mRNA half-life. However, the effects of SB203580 on G-CSF expression in macrophages are unclear. In this study, we investigated the effects of SB203580 on G-CSF expression in LPSstimulated macrophages. Surprisingly, we found that SB203580 increases both G-CSF mRNA and protein levels in LPS-stimulated RAW264.7 macrophages and mouse bone marrow derived macrophages (BMDM). In contrast, levels of IL-1β mRNA were lower in SB203580/LPS treated cells compared to that in LPS treated cells. To test if increase of G-CSF mRNA in the SB203580/LPS treated cells is resulted from the increase of transcription, we investigated the effect of SB203580 and LPS on G-CSF promoter activity by luciferase reporter assay and evaluated transcription rate by nuclear run on assay. The results showed that luciferase activity and transcription rate are both upregulated by LPS, while SB203580 decreased luciferase activity without affecting transcription rate. We then tested if G-CSF mRNA stability is enhanced by SB203580, LPS or SB203580/LPS treated cells were further treated with actinomycin D, mRNA were isolated at 0, 30, 60, 90 and 120 min after actinomycin D treatment, and the levels of G-CSF mRNA were determined by Northern blotting. The results show that G-CSF mRNA is more stable in SB203580/LPS treated cells than that in cells treated with LPS only. To investigate the mechanisms that mediate the increase of G-CSF mRNA stability in response to SB203580/LPS treatment, the G-CSF 5’UTR, 3’UTR and the G-CSF coding regions were cloned into the 5’or 3’ end of luciferase gene in the pGL3-control vector to produce pGL3-G-CSF 5’UTR-Luc, pGL3-Luc-G-CSF 3’UTR, pGL3-Luc-G-CSF(+35/+372) and pGL3-Luc-G-CSF(+340/+720) plasmids. These plasmids were then transfected into RAW264.7 cells and the luciferase activities were determined after treating with SB203580 and/or LPS. We found that the cells transfected with pGL3-Luc-G-CSF 3’UTR and pGL3-G-CSF 5’UTR-Luc plasmid expressed very low luciferase activity, and the treatment with SB203580 did not increase the luciferase activity in LPS-stimulated cells. However, cells transfected with pGL3-Luc-G-CSF(+35/+372) or pGL3-Luc-G-CSF(+340/+720) express luciferase activities similar to that of cells transfected with pGL3-control, and the treatment of SB203580/LPS resulted in slightly higher luciferase activities. The data suggest that 3’UTR and 5’UTR mediated mRNA de-stabilization while the coding region may contain sequences that increase mRNA stability. However, in siRNA mediated p38α knockdown cells, LPS-stimulation did not result in the increase of G-CSF mRNA compared with control siRNA transfected cells. Taken together, our data show that p38 inhibitor SB203580 induced expression of both G-CSF mRNA and protein, which is resulted from the increase, at least in part, of G-CSF mRNA stability; however, the effect may not due to the inhibition of p38 in macrophages