Use of yeast expression technology to study G protein-coupled receptor function

为了了解GPCR如何在分子水平上发挥作用，重要的是要确定对配体结合，受体活化和生产性受体/G蛋白偶联至关重要的氨基酸。我们最近描述了一种新的实验策略，其允许快速鉴定M3毒蕈碱受体（M3 R）（原型I类GPCR）中的功能关键氨基酸（Li等人，Nat. Methods 4，169-74，2007）。该方法涉及整个M3 R编码序列的随机诱变（除了第三胞内环的中心部分），然后进行酵母遗传筛选，该筛选允许回收含有破坏受体功能的单点突变的突变M3 R。这种方法明显不同于以前使用的酵母遗传筛选，后者被设计用于恢复保留功能活性的突变受体。 我们通过易错PCR使用最大化单核苷酸变化发生的条件产生突变体M3 R文库。所有突变体M3 R在MPY 578 q5菌株中表达，该菌株表达嵌合酵母/哺乳动物G蛋白α亚基并含有几种其他修饰，这些修饰允许通过功能性M3 R有效激活酵母MAP激酶/信息素途径。我们利用了MPY 578 q5菌株含有信息素敏感的FUS 2-CAN 1报告基因的事实。因此，只有表达功能失活突变M3 R的酵母克隆在同时存在细胞毒素刀豆氨酸（该试剂需要Can 1 p的表达才能进入酵母细胞）和毒蕈碱激动剂卡巴胆碱的情况下才能生长。 将所有突变引入含有C-末端EGFP标签的M3 R的修饰版本中。结果，在刀豆氨酸/卡巴胆碱选择程序中存活但未能显示出稳健GFP荧光的酵母克隆在筛选期间早期被消除。这些酵母克隆含有由无义或移码突变引起的截短受体，或以低水平表达的突变M3 R。因此，这种策略使我们能够回收含有错义突变的全长突变体M3 R，这些错义突变不会导致蛋白质稳定性的大幅降低。 这种筛选策略的应用导致了174个单点突变的恢复，这些突变消除了酵母中的M3 R信号传导。对转染的哺乳动物（COS-7）细胞的功能研究表明，绝大多数这些突变型M3 R（174个中的159个）不能或显著损害其诱导卡巴胆碱介导的细胞内钙水平升高的能力。 先前的定点诱变研究已经鉴定了M3 R和其他毒蕈碱受体亚型中的许多氨基酸，这些氨基酸在毒蕈碱受体功能中起关键作用。在本研究中也鉴定了以前鉴定的这些功能关键氨基酸中的绝大多数。这些包括例如在I类GPCR中高度保守的氨基酸和已知对乙酰胆碱结合重要的残基。总之，这些观察结果令人信服地验证了这种新的酵母筛选策略的有用性。 我们还发现了许多新的点突变，这些突变干扰了以前没有描述过的M3 R功能。有趣的是，大多数这些取代涉及位于跨膜结构域I和II（TM I和II），两个M3 R区域，在过去没有进行系统的定点诱变研究的氨基酸。分子模拟研究表明，大多数的诱变TM I和II残基的项目到内部的受体蛋白质或对其他TM螺旋。我们假设这些残基是分子内氢键和/或疏水相互作用网络的一部分，预测对M3 R活化至关重要。 最近的研究表明，I类GPCR的第二个细胞外环（o2环）与跨膜受体核心密切接触。在过去，o 2环在激动剂依赖性GPCR激活中的潜在作用尚未被系统地研究。使用Li等人（Nat. Methods 4，169-74，2007）描述的类似实验策略，我们发现M3 R的o2环中的几个残基在调节激动剂诱导的受体活化的效率中起重要作用（Scarselli等人，282，7385-96，2007）。一般而言，这些残基的突变修饰对激动剂结合亲和力几乎没有影响。我们的研究结果是一致的模型，其中多个O2环残基参与稳定的活性状态的M3 R。 鉴于酵母表达技术的最新进展，应该可以在酵母中以可偶联形式表达大多数GPCR。因此，我们开发的新的实验策略（Nat. Methods 4，169-74，2007）应该适用于许多或大多数哺乳动物GPCR。由于已知GPCR代表药物治疗的优良靶点，因此该策略应具有广泛的普遍相关性和相当大的临床意义。

To understand how GPCRs function at the molecular level, it is important to identify the amino acids that are critical for ligand binding, receptor activation, and productive receptor/G protein coupling. We recently described a novel experimental strategy that allows the rapid identification of functionally critical amino acids in the M3 muscarinic receptor (M3R), a prototypic class I GPCR (Li et al., Nat. Methods 4, 169-74, 2007). This method involves random mutagenesis of the entire M3R coding sequence (except for the central portion of the third intracellular loop), followed by a yeast genetic screen that allows the recovery of mutant M3Rs containing single point mutations that disrupt receptor function. This approach clearly differs from previously used yeast genetic screens which were designed to recover mutant receptors that retain functional activity. We generated mutant M3R libraries via error-prone PCR using conditions to maximize the occurrence of single nucleotide changes. All mutant M3Rs were expressed in the MPY578q5 strain which expresses a chimeric yeast/mammalian G protein alpha subunit and contains several other modifications that allow efficient activation of the yeast MAP kinase/pheromone pathway by functional M3Rs. We took advantage of the fact that the MPY578q5 strain contains the pheromone-sensitive FUS2-CAN1 reporter gene. As a result, only yeast clones expressing functionally inactive mutant M3Rs will grow in the simultaneous presence of the cytotoxin, canavanine (this agent requires the expression of Can1p to enter the yeast cell), and the muscarinic agonist, carbachol. All mutations were introduced into a modified version of the M3R that contained a C-terminal EGFP tag. As a result, yeast clones that survived the canavanine/carbachol selection procedure but failed to display robust GFP fluorescence were eliminated early during the screen. These yeast clones contained either truncated receptors caused by nonsense or frameshift mutations, or mutant M3Rs which were expressed at low levels. Thus, this strategy allowed us to recover full-length mutant M3Rs containing missense mutations that did not lead to major reductions in protein stability. Application of this screening strategy resulted in the recovery of 174 single point mutations that abolished M3R signaling in yeast. Functional studies with transfected mammalian (COS-7) cells showed that the vast majority of these mutant M3Rs (159 out of 174) were either unable or significantly impaired in their ability to induce carbachol-mediated increases in intracellular calcium levels. Previous site-directed mutagenesis studies have identified many amino acids in the M3R and other muscarinic receptor subtypes that play critical roles in muscarinic receptor function. The vast majority of these functionally critical amino acids identified previously were also identified in the present study. These include, for example, amino acids that are highly conserved among class I GPCR and residues known to be important for acetylcholine binding. Taken together, these observations convincingly validated the usefulness of this novel yeast screening strategy. We also recovered many novel point mutations that interfered with M3R function that had not been described previously. Interestingly, the majority of these substitutions involved amino acids located within transmembrane domains I and II (TM I and II), two M3R regions that have not been subjected to systematic site-directed mutagenesis studies in the past. Molecular modeling studies suggested that most of the mutagenized TM I and II residues project into the interior of the receptor protein or towards other TM helices. We hypothesize that these residues are part of network of intramolecular H-bonds and/or hydrophobic interactions predicted to be critical for M3R activation. Recent studies suggest that the second extracellular loop (o2 loop) of class I GPCRs is in close contact with the transmembrane receptor core. In the past, the potential role of the o2 loop in agonist-dependent GPCR activation has not been studied systematically. Using a similar experimental strategy as described by Li et al. (Nat. Methods 4, 169-74, 2007), we found that several residues in the o2 loop of the M3R play important roles in regulating the efficiency of agonist-induced receptor activation (Scarselli et al., J. Biol. Chem. 282, 7385-96, 2007). In general, mutational modification of these residues had little effect on agonist binding affinities. Our findings are consistent with a model in which multiple o2 loop residues are involved in stabilizing the active state of the M3R. Given recent advances in yeast expression technology, it should be possible to express most GPCRs in yeast in a coupling-competent form. Thus, the novel experimental strategy that we developed (Nat. Methods 4, 169-74, 2007) should be applicable to many or most mammalian GPCRs. Since GPCRs are known to represent excellent targets for drug therapy, this strategy should be of broad general relevance and considerable clinical interest.