調控蛋白質作用中心的機要生理化學原則：半胱胺酸與金屬活性中心

Abstract

蛋白質的活性在細胞與分子層級皆受到嚴密的調控，分子層面牽涉調控蛋白質局部的環境以利蛋白質效能的最佳化，目前對於如何調控蛋白作用點局部環境的生化準則，相關研究仍十分缺乏，因此我們試圖瞭解具有重要生化功能的氨基酸如何相互作用達到局部蛋白環境的最佳化。我們以半胱胺酸(cysteines)和鋅(Zn2+)及鈉(Na+)金屬蛋白質為研究系統，利用量子化學與模擬周邊環境作用的高階計算方法進行研究，計算結果並輔以實驗可得的數據進行分析與比較。 半胱胺酸的反應性和蛋白質環境息息相關，其中一個至關重要的調控因子便是氫鍵，因此我們對最有可能和半胱胺酸形成氫鍵的氨基酸進行分析，研究結果顯示半胱胺酸形成氫鍵的選擇性和強度主要受到周圍環境和與其產生氫鍵的氨基酸性質等綜合因素的影響。藉由提供適當可與之形成氫鍵的不同氨基酸與環境介電常數，蛋白質可以調控氫鍵的強度並連帶操控硫醇化合物(thiolate)的穩定性，而這也會影響半胱胺酸的化學親核性與反應性。利用已知晶體結構對半胱胺酸和其蛋白質環境進行統計分析，我們發現半胱胺酸在蛋白質中形成氫鍵的對象和理論計算的預測吻合。 在蛋白質中半胱胺酸喜好和鋅金屬鍵結，然而由於半胱胺酸和軟性過渡金屬的高度結合力，鋅很容易被有毒的鎘(Cd2+)金屬取代。蛋白質通過和直接和間接金屬配位基所形成的氫鍵來調控並且保護金屬蛋白或者強化和金屬行成化學鍵的能力。然而，我們對各類金屬蛋白中，不同金屬偏好形成氫鍵的對象氨基酸仍缺乏瞭解。我們是第一個以系統性方法闡述蛋白質中調控與鋅及鎘金屬形成氫鍵的氨基酸的準則。金屬的直接配位基與周圍環境可以決定喜好形成氫鍵的對象氨基酸，也就是金屬的間接配位基。在鋅與鎘的取代反應中，兩金屬離子的半徑大小差異對他們蛋白質環境的選擇性有著相似及全然不同的影響。我們的結論是電中性與配位環境可變更的鋅離子反應中心如果配以中性的小型氨基酸配位基便極容易被有毒的鎘金屬取代，因為鎘金屬會使蛋白質原有的結構變形進而改變或破壞蛋白質的功能。 和鋅蛋白類似，鈉(Na+)金屬蛋白在細胞中也扮演許多重要的角色，雖然鈉與鎂(Mg2+)離子在細胞質中的濃度相似，這兩種離子蛋白如何區分鈉、鎂和環境中其他離子，還有蛋白質調控鈉與鎂離子選擇性的物理原理都是未知的領域。我們的研究發現鈉蛋白對於鈉離子優於鎂的選擇性主要是基於金屬離子大小、價性和電子接收能力的差異性，這些結果和目前收集到的實驗數據相符。 我們接著研究鈉離子和非生物起源的鋰(Li+)離子在神經傳導蛋白與G受體蛋白中的競爭性，這兩種蛋白質是目前精神病徵與藥物依賴方面為人熟知的主要藥物投遞目標，特別是鋰離子一直是治療多種精神疾病例如憂鬱症等的第一線藥物。佈滿蛋白質配位基的鈉金屬活性中心如果僅能維持原來構型而缺乏結構的變化性就不會受到鋰離子的影響而被取代，相對的在具有結構變化度或者僅有一、兩個配位基的金屬活性中心便容易被鋰離子取代。在G受體蛋白中，當鋰離子取代水分子不易進入的鈉離子金屬中心與天門冬胺酸(aspartic acid)和絲胺酸(serine)鍵結時，鋰離子可以藉由抑制構型改變來穩定處於不被活化狀態的受體蛋白質，如此可以使躁鬱症病人過度被活化的G蛋白活性降低。 從我們的模擬系統中，我們證實了支配主要蛋白質局部環境最佳化與調控功能的生理化學原則，這些結果可以廣泛得運用在蛋白質設計、理性藥物設計以及生物感應元件的發展。這篇論文中更指出，未來深入探討調控此類問題生理化學準則的重要性，並且概述值得進一步研究的相關問題。

Protein activity in the cell is tightly regulated on both cellular and molecular levels. Molecular level involves regulation through creating a local protein environment around a functional site organized for an optimal protein activity. Currently, general physicochemical principles of regulation of local environment of functional sites in proteins remain elusive and in this study, we tried to answer the question of how can interactions of the functionally important residues favor the optimal organization of the local protein environment. As examples of such functional systems, we have chosen free cysteines, zinc and sodium-binding proteins. To address the posed question we modeled the studied functional systems together with their interactions explicitly and implicit solvent with the quantum chemical density functional (QC/DFT) approach. The computational results were complemented with analysis and survey of the available experimental data. Cysteine (Cys) reactivity depends on the protein environment and one of the key contributors to its modulation is the HB interactions. Thus, we assessed the most preferred HB partners of free Cys. As a result, we show that the strength and selectivity of the HB interactions with free Cys ultimately depend on the combination of the extent of HB complex solvation and the properties of Cys HB partner. In this manner, through providing various HBs partners and dielectric environment the protein of interest can modulate the HB strength and consequently the degree of thiolate stabilization, which in turn would affect the cysteine's nucleophilicity/reactivity. Statistical analysis of free Cys and their local protein environment in the reported crystal structures revealed that in the proteins the HB interactions with Cys are formed according to our predicted HB preferences. In proteins, cysteines are preferred first-shell ligands for zinc (Zn2+), and zinc is readily displaced by toxic cadmium (Cd2+) due to the high affinity of the later for cysteines. Proteins through providing HB interactions between the first- and second-shell ligands regulate and stabilize/protect the metal complex or enhance metal binding properties. Yet, knowledge of the preferred HB partners of metal ligands in different metalloproteins is lacking. We were the first to systematically unravel the principles governing the preferred hydrogen-bonding partners of metal ligands on the example of zinc- and cadmium-thiolate (Cys-) complexes in proteins. It appears that the nature of metal-binding first-shell residues and solvent exposure of the complex determine the HB preferences of the second shell ligands. Upon Zn2+ to Cd2+ substitution, the size difference of the two cations yielded similarities and differences in their protein environment preferences. We conclude that neutral and flexible Zn2+-sites lined by small neutral amino acid (aa) sidechains are more vulnerable for toxic Cd2+ effects as it could distort the native protein structure and thereby alter/abolish the protein function. Similarly to zinc proteins, sodium-binding proteins play diverse important roles in the cell. However the concentration of Na+ and Mg2+ cations are present in comparable amounts in the cytosol, but it is not clear how Na+ and Mg2+ binding sites discriminate the native cation among non-cognate ones from the surrounding milieu and the physical basis governing the selectivity for Na+ over Mg2+. The results, which are consistent with available experimental data, reveal that in proteins, the selectivity for Na+ over Mg2+ in sodium-binding sites stem mainly from the size, charge, and charge-accepting ability differences between Na+ and Mg2+. Next, we studied the competition between native Na+ and abiogenic Li+ in sodium neurotransporters and G-protein coupled receptors (GPCRs). Neurotransmitter transporters and receptors are well-known drug targets for psychiatric disorders and addictive behavior and lithium has been used as a first-line treatment of the number of psychiatric disorders (e.g. depression). We show that rigid Na+-sites that are crowded with multiple protein ligands are well-protected against Li+ attack, but their flexible counterparts or buried Na+-sites containing only one or two protein ligands are vulnerable to Li+ substitution. By displacing Na+ bound to an Asp- and a Ser in the solvent-inaccessible metal-binding sites of the A2AAR adenosine and the 1AR adrenergic GPCRs, Li+ could stabilize the receptor's inactive state by prohibiting conformational changes to the active state, thus leading to decreased G-protein activity, which are hyperactive in bipolar patients. In our studied model systems, we have demonstrated how key physicochemical factors govern the formation of the optimal local environments around functional sites allowing proteins to regulate their functions. Our results can be broadly applied to artificial protein design, rational drug design, and development of biosensors. We underscore the importance of further investigations of physicochemical local environments of the functional sites and their role in the regulation of protein functions and outline several problems for future investigations.