BCL10磷酸化位置與MALT1誘導切割位確認的研究

Abstract

BCL10與MALT1均是淋巴細胞抗原受體BCR/TCR至NF-κB活化過程中重要的訊息傳遞分子，過去研究發現NF-κB活化作用經常伴隨BCL10磷酸化修飾共同出現。然而，BCL10磷酸化具有何種生理意義目前仍不清楚。因此，我們期望藉由找出BCL10磷酸化位置，以對於BCL10磷酸化有所瞭解。此外，共同表現MALT1與BCL10的研究中，我們發現BCL10會被誘導切割。因此，希望找出BCL10受MALT1誘導之切割位。 BCL10具有233個胺基酸，包含N端CARD domain，可藉此與其他具有CARD domain的分子結合；C端Serine/Threonine rich domain，根據過去研究推測，可能是磷酸化的主要標的區域。大量表現BCL10時會發生磷酸化現象、NF-κB活化，且在顯微鏡下可見絲狀構造產生。當BCL10第41個高度保留之胺基酸Leucine突變，則CARD domain失去與其他分子結合的能力，且磷酸化現象消失、蛋白質均勻分佈於細胞質和細胞核，NF-κB不活化。N端(1~114 a.a.)或C端(114~233 a.a.) BCL10均無高度磷酸化現象，顯示BCL10磷酸化需同時具有CARD及Serine/Threonine rich domain。然而，CARD domain正常之N端BCL10仍可活化NF-κB，並能產生絲狀構造，顯示BCL10磷酸化並非活化NF-κB的必要因素。共同轉染BCL10與可抑制NF-κB活化的Dominant Negative-IκB質體至293T細胞，發現BCL10磷酸化程度減弱。表示BCL10磷酸化雖不是NF-κB活化的必要因素，但會受到NF-κB活化作用影響。針對BCL10可能磷酸化區域進行刪除或點突變作用，構築BCL10 deletion constructs D1(△134~150 a.a.)、D2(△185 a.a.)、D3(△218~231 a.a.)、D4(△134~150 & 218~231 a.a.)，及site mutated BCL10 constructs S7A、S48A、S60,61A、S85A。這些表現質體仍具有活化NF-κB、產生絲狀構造及磷酸化修飾的能力，只是不同mutants之間磷酸化程度略有差異，由上述結果推測BCL10可能具有多重磷酸化修飾。此外，我們發現BCL10表現量越少則磷酸化現象越不明顯。 以TPA/ionomycin刺激Jurkat細胞，內生性BCL10似有被切割現象。據報導MALT1有paracaspase特性且能與BCL10產生結合作用。為檢測MALT1是否具有誘導BCL10被切割的可能性，我們將不同的BCL10 constructs與MALT1共同表現，以西方墨點法分析切割情形。結果顯示除了D3和D4無法被誘導切割之外，其餘mutant BCL10均可被誘導切割，推測D3與D4共同被刪除之區域(218~231 a.a.)具有可能的MALT1誘導切割位。進一步針對此區域構築BCL10 deletion constructs C1 (△218~224 a.a.)、C2 (△219~225 a.a.)、C3 (△220~226 a.a.)、C4 (△221~227 a.a.)、C5 (△222~228 a.a.)、C6 (△223~229 a.a.)、C7 (△224~230 a.a.)、C8 (△225~231 a.a.)來研究切割位置。結果僅C1可被誘導切割，顯示BCL10 Leucine 225可能是MALT1誘導切割位。而由誘導切割後之產物分子量大於原本BCL10分子量的現象看來，在切割產物上似乎具有修飾作用，經由去磷酸酵素實驗證實是磷酸化修飾的結果。由於切割作用與BCL10磷酸化形式比例減少的現象常伴隨出現，並且BCL10L41RGFP與C’BCL10GFP均無法被有效切割，表示磷酸化的BCL10才是較好的MALT1誘導切割受質。 MALT1具有一個Death Domain、兩個Immunoglobulin like domain及caspase like domain (CLD)。我們利用五種MALT1表現質體： MALT1△N (127~824 a.a.)、MALT1 1~548 a.a.、MALT1△498~548 a.a.、MALT1△199~498 a.a. 與IAP2-MALT1 (MALT lymphoma基因轉位之融合蛋白質)檢視哪些區域負責於誘導切割作用發生。結果顯示僅MALT1△N與IAP2-MALT1保有誘導切割能力，MALT1 1~548 a.a.、MALT1△498~548 a.a.以及MALT1△199~498 a.a.則無法誘導切割。表示MALT1需同時具有CLD與其C-端區域才能執行誘導切割作用。

Both BCL10 and MALT1 function as signaling regulators in antigen receptor-mediated NF-κB activation. The activation of NF-κB is often associated with BCL10 phosphorylation, but the physiological significance of BCL10 phosphorylation is still unknown. In this study, we would like to identify the phosphorylation site(s) of BCL10 and try to unravel the myth of BCL10 phosphorylation. Besides, we found that BCL10 could be cleaved when co-expressed with MALT1. Therefore, we would like to locate the MALT1-induced cleavage site of BCL10. BCL10 interacts with several CARD-containing proteins via its N-terminal CARD (caspase recruitment domain), and its C-terminal Serine/Threonine rich domain is probably the target region of phosphorylation. Overexpression of BCL10 induced BCL10 phosphorylation, NF-κB activation and the appearance of filamentous structure under microscope. BCL10L41R, a CARD mutated BCL10, lost all these BCL10-induced abilities. Neither N’BCL10GFP (1~114 a.a.) nor C’BCL10GFP (114~233 a.a.) was phosphorylated, indicating that both CARD and Serine/Threonine rich domain are required for BCL10 phosphorylation. Though not phosphorylated, N’BCL10GFP retained the ability to activate NF-κB, suggesting that BCL10 phosphorylation was not essential for NF-κB activation. The level of BCL10 phosphorylation was diminished by dominant-negative IκB through blocking NF-κB activation. BCL10 deletion constructs：D1(△134~150 a.a.),D2(△185 a.a.),D3(△218~231 a.a.),D4(△134~150 & 218~231 a.a.) and site mutated BCL10 constructs：S7A,S48A,S60,61A,S85A were generated to investigate the phosphorylation site(s) of BCL10. All the mutants maintained abilities to activate NF-κB, form filamentous structure, and be phosphorylated, though the ratio of phosphorylated form to un-phosphorylated form of these constructs were more or less different. These results implied that there were multiple phosphorylation sites on BCL10 molecule. Moreover, the phosphorylation level of BCL10 correlated to its expression level. Endogenous BCL10 was observed to be processed into a shorter molecule in Jurkat cells treated with TPA/ionomycin. MALT1 was reported to be a paracaspase and able to interact with BCL10. We then tested the possibility as whether MALT1 could mediate the cleavage. With the co-expression of individual BCL10 constructs and MALT1, cell lysates were analyzed by western blotting. Except D3 and D4, BCL10 as well as other BCL10 mutants were shown to be cleaved. Therefore, the common deleted region (218~231 a.a.) of D3 and D4 contained the probable cleavage site. Serial BCL10 deletion constructs：C1 (△218~224 a.a.), C2 (△219~225 a.a.), C3 (△220~226 a.a.), C4 (△221~227 a.a.), C5 (△222~228 a.a.), C6 (△223~229 a.a.), C7 (△224~230 a.a.) and C8 (△225~231 a.a.) were created to pinpoint the precise cleavage site. Since C1 was the only one to be cleaved, leucine 225 of BCL10 was predicted to be the best cleavage site candidate. The cleaved products had M.W. larger than authentic BCL10, indicating possible post-translational modification. In fact, CIAP (Calf Intestine Alkaline Phosphatase) treatment relieved the slower migrating property of the cleaved products. The cleaved products were often observed to be accompanied with the diminished level of phosphorylated BCL10. Meanwhile, BCL10L41RGFP and C’BCL10GFP were not good substrates for cleavage. These results supported the idea that phosphorylated BCL10 was a better substrate for MALT1-induced cleavage. MALT1 contains a death domain, two Immunoglobulin like domain and a caspase like domain. To map the domain responsible for the cleavage, five MALT1 constructs：MALT1△N (127~824 a.a.), MALT1 1~548 a.a., MALT1△498~548 a.a., MALT1△199~498 a.a. and IAP2-MALT1 (the fusion protein of IAP2 and MALT1 identified from MALT lymphoma) were tested. While MALT1△N and IAP2-MALT1 retained their abilities as wild-type MALT1, MALT1 1~548 a.a., MALT1△498~548 a.a. and MALT1△199~498 a.a. lost their abilities to induce BCL10 cleavage. Therefore, both CLD and C-terminal region of MALT1 were required for mediating the cleavage.