乙醯化之蛋白質轉譯後修飾經由調節新陳代謝相關酵素控制存活壽命

Abstract

在本論文中，我們首度利用真核生物的完整蛋白質體篩檢平台，找到酵母菌唯一的存活所不可或缺的賴氨酸乙醯基轉移酶 (lysine acetyltransferase) 所作用的下游非組蛋白受質 (non-histone protein substrates)，以及第一批受質的乙醯化位點 (acetylation sites)。 在第一部分的研究中，我們利用蛋白質體微陣列晶片 (proteome microarray)，找到NuA4蛋白質複合體 (其催化次單元為Esa1) 這個酵母菌唯一與存活有關的重要賴氨酸乙醯基轉移酶 (essential lysine acetyltransferase) 的十三個乙醯化受質，並發現他們大多為新陳代謝以及細胞因應能量壓力有關的重要蛋白質。在這些受質之中，phosphoenolpyruvate carboxykinase (Pck1) 為位於細胞質內，一個葡萄糖新生 (gluconeogenesis) 作用的關鍵酵素。此一發現暗示NuA4有除了調控細胞核內染色質有關的細胞過程以外，也控制細胞核外、細胞質內的重要新陳代謝相關功能。利用串聯質譜儀 (tandem mass spectrometry) 的技術，我們找到Pck1被乙醯化之賴氨酸位點 (lysine 514, or K514)。利用生化技術，我們進一步確認Pck1的賴氨酸去乙醯酶 (lysine deacetylase) 為Sir2。Pck1的K514乙醯化會影響其催化葡萄糖新生的酵素活性：當把K514突變為精氨酸 (arginine, or pck1-K514R) ，以模擬去乙醯化狀態之Pck1 時，Pck1活性幾乎完全消失。此外，帶有此突變的酵母菌 pck1-K514R，在乙醇培養基上的存活能力變差。相反的，當把K514突變為麩醯氨（glutamine, or Q）時，模擬強制乙醯化狀態之pck1-K514Q，不僅對於高濃度的乙醇有較佳的耐受力，也會挽救esa1溫度突變菌株在乙醇培養基上存活變差的現象。顯示Pck1為NuA4下游，調控使用乙醇為替代碳水化合物來源時，絕對必須的葡萄糖新生作用一個很重要的受質。我們也進一步證實，Pck1對於酵母菌在能量限制狀態下，時間存活壽命 (chronologic life span) 的延長，扮演關鍵的角色。 在第二部份的研究中，我們發現NuA4 這個重要的賴氨酸乙醯基轉移酶，會乙醯化另一個非組蛋白叫做Sip2。Sip2是Snf1 蛋白質複合體的三個beta次單元其中之一，而Snf1蛋白質複合體為酵母菌的AMP-activated protein kinase (AMPK) 同源基因 (homologue)。我們的研究發現，Sip2蛋白質之乙醯化程度，會隨酵母菌老化而減少。利用串聯質譜儀，我們找到四個Sip2蛋白質的乙醯化位點：賴氨酸 (lysine, K) K12、16、17 以及256。我們也證明Sip2 之乙醯化，是由 互相拮抗的賴氨酸乙醯基轉移酶 NuA4 ，以及賴氨酸去乙醯酶 Rpd3 所共同控制。模擬乙醯化狀態的SIP2突變 (sip2-4KQ)，會顯著延長酵母菌的複製存活壽命；相反的，模擬去乙醯化狀態的SIP2突變 (sip2-4KR)，與敲毀SIP2基因相同，都會顯著縮短酵母菌的複製存活壽命。Sip2 之乙醯化，會增加其與Snf1催化次單元的結合能力與交互作用。Sip2-Snf1之交互作用會抑制Snf1的催化活性，因而減少其下游標的基因--Sch9 (酵母菌之Akt/S6K同源基因) 的磷酸化及活化，最終導致生長減緩、卻可延長酵母菌的複製存活壽命 (replicative life span)。模擬Sip2 乙醯化狀態的基因突變株 (sip2-4KQ)，其所含有之壓力指標物—trehalose的濃度較低，且對於與老化有關的氧化壓力具有較佳的耐受性。我們也進一步證實Sip2乙醯化的抑制生長及抗老化效果，是獨立於外在營養之可近性，以及TORC1的活性。我們認為此一新發現之隨細胞老化而逐漸減少的蛋白質乙醯化、與逐漸增強的磷酸化訊息傳遞路徑，藉由調控重要的下游基因Sch9 之磷酸化與活性，會經由控制細胞內在環境的老化壓力，從而影響細胞的複製存活壽命。 我們的研究，首度證明賴氨酸乙醯化的非組蛋白受質，在時間存活壽命與複製存活壽命所扮演的重要角色。未來希望可以將此結果應用於人類與老化有關的新陳代謝疾病與癌症之相關研究。

In this essay we report the use of whole proteome screen in eukaryotes to discover the first non-histone substrates of the only essential lysine acetyltransferase, and the first non-chromatin acetylation sites in yeast. In the first study, we used the yeast proteome microarray to identify 13 in vivo substrates of the only essential lysine acetyltransferase NuA4 complex (containing the Esa1 catalytic subunit) in yeast. Many of the in vivo substrates are metabolic enzymes and stress-response proteins, including phosphoenolpyruvate carboxykinase (Pck1), a well-characterized enzyme catalyzing the rate-limiting step in gluconeogenesis in the cytoplasm, indicating a surprising extranuclear function of NuA4 in regulating metabolism besides its well-studied regulation of chromatin-related processes. Using mass spectrometry, we identified lysine residue 514 (K514) as the Esa1-dependent acetylation site in Pck1, and further showed the same site was deacetylated by lysine deacetylase Sir2. Moreover, we found that mutations of K514 affected enzymatic activity of Pck1 to conduct gluconeogenesis and hence the ability of cells to grow on non-fermentable carbon sources such as ethanol. When K514 was mutated to arginine (pck1-K514R) to abolish acetylation at this critical site, the enzymatic activity of purified Pck1 was markedly decreased in vitro, and the cells lost the ability to grow in ethanol. By contrast, substituting K514 with glutamine (pck1-K514Q, a mutation mimicking constitutive acetylation) could almost completely rescue the lethality of mutants with defective Esa1 function in ethanol. These results suggested that K514 is the only lysine residue of Pck1 targeted by Esa1 and Sir2. Interestingly, both pck1-K514Q and sir2∆ mutants enhance the viability of cells on high concentration of ethanol. Loss of Pck1 activity by either deletion of the encoding gene or by mutation of K514 to arginine blocked the extension of chronological life span under calorie restriction. Furthermore, this activating acetylation might be conserved in mammalian system since human Pck1 could rescue the lethality of yeast cells lacking PCK1 in ethanol, despite a low sequence homology, and hPck1 acetylation and glucose production was dependent on TIP60 in human hepatocellular carcinoma (HepG2) cells. In summary, we have found novel extracellular functions of yeast NuA4 complex in regulating gluconeogenesis and chronological life span. Aging is a plastic phenotype determined partially by the cellular metabolism and energy expenditure. Our second study describes a novel regulatory cascade mediated by acetylation and phosphorylation that modulates cellular metabolism, growth and replicative life span in yeast. At the top of the pathway is the essential lysine acetyltransferase NuA4, which acetylates a regulatory subunit of the Snf1 complex (yeast AMPK), called Sip2. This acetylation blocks kinase action by stabilizing Sip2-Snf1 protein interaction; in response to deacetylation, Snf1 is released from Sip2 inhibition and then activates a life span shortening pathway by phosphorylating Sch9. Snf1 kinase is a key regulator of energy homeostasis required for transcription of glucose-repressed genes and certain stress-response genes. A previously unsolved mystery is why Snf1 activity increases with aging and exhibits negative effects on life span extension. We show that Sch9, the yeast homologue of mammalian Akt/S6K, a known life span antagonist, is phosphorylated and activated by Snf1. Acetylation of Sip2 enhances physical interaction with Snf1, antagonizes its catalytic activity and suppresses detrimental effects of Snf1 on life span extension. Decreased Sip2 acetylation during aging enhances activity of Snf1 and the downstream Sch9 kinase. We further demonstrate that Sip2 acetylation decreases intracellular trehalose level, a stress indicator, and increases resistance to aging-associated oxidative stress. Whether calorie restriction is the only pathway and approach for life span extension is unclear. Previous reports suggested that calorie restriction might lead to unwelcome health concerns in humans, especially elderly and non-obese subjects. Our study provides evidence of a potential “intrinsic aging pathway” mediated by an acetylation-phosphorylation cascade that is largely unresponsive to calorie restriction. Although calorie restriction increases the life span of non-acetylatable Sip2 mutants, or when SIP2 is deleted (both mimicking the aging status); the benefit of glucose limitation to constitutively acetylated Sip2 mutants is limited, indicating a partially overlapping common downstream pathway for calorie restriction and intrinsic aging. Our study confirmed the relationship between non-histone protein acetylation and both chronological and replicative life span in yeast. We hope that in the future, the results can be applied to aging-related diseases, such as diabetes mellitus and cancer, in higher organisms and also humans.