A central strategy for converting natural products into fluorescent probes
A central strategy for converting natural products into fluorescent probes
复制标题
DOI：
10.1002/cbic.200500466
复制
发表时间：
2006-03-01
期刊：
影响因子：
3.2
通讯作者：
La Clair, JJ
中科院分区：
文献类型：
--
作者：
Alexander, MD;Burkart, MD;La Clair, JJ
在Wieland用异硫氰酸荧光素标记鬼笔环肽的研究近40年后，[1]荧光修饰已经成为研究天然产物生物学的重要工具。[2-5]自从维兰德的努力，有一个深刻的增加，在标签的数量。这种增加导致了天然产品探针的复杂纲要的产生，正如从标签A-D的附属物到天然产品a-d产生Aa-Dd所说明的那样（图1A）。由此产生的纲要Aa-Dd现在包含三个不同变量的修改：标签，链接和天然产物。虽然标记和接头的变化可能在活性优化中有用，但Aa-Dd的扩展多样性深刻地抑制了进行比较分析的能力。通过使用单个标记和一组共同的接头，减少了天然产物探针集合内显示的变量的数量。因此，从改良的探针组Da-Dd收集的数据被均质化（图1B）。我们的研究开始于确定一种从单一染料标记多种天然产物的途径。在选择最佳染料时，我们考虑了以下标准的组合，例如良好的生物物理性质、足够的细胞摄取、缺乏亚细胞定位和缺乏内在生物活性。对于小分子缀合，碳菁染料和罗丹明染料，包括罗丹明123和罗丹明B，被摄取到线粒体中，而其他常见标记，如花菁染料或BODIPY染料分别定位在细胞核和内质网（ER）中。7-二甲基氨基香豆素-4-乙酸乙酯（1）[6-7]因其缺乏生物活性而被选为这些研究的中心染料。香豆素1很容易被细胞吸收，但是，它不定位，可以通过洗涤细胞去除。此外，它的水溶性，[8]小尺寸，[9]可用性，[10]光稳定性，[11]和生物物理性质（lex= 370 nm，e= 22 000 M·cm-1，lem= 459 nm，F= 0.1-0.4）有利于合成应用的发展，同时为各种生物物理实验提供可行的荧光。[12]香豆素1通过使用该团队成员开发的程序转化为一组官能化标记2-9。[7-8]通过使用该途径，常规地由间-（N，N）-二甲基氨基苯酚和1，3-丙酮二羧酸二乙酯制备10-100 g批次的1，并转化为酸2。由酯1或酸2制备一组标记物。如方案1所示，将2与甘氨酸叔丁酯偶联，然后用TFA脱保护并用N-羟基琥珀酰亚胺酯化，得到酯3，总产率为25%。可比较的酰胺偶联在合成标记4-6和8-9中也是有效的（方案1）。或者，标记物3-9可以通过伯胺与酯1的硫醇辅助加成来制备，如半胱胺与1反应得到7所示。[13-14]加入5-20摩尔%的硫酚钠、乙硫醇钠或IV族金属醇盐[15]可用于促进该过程。有了标记2-9，我们的努力转向筛选反应条件，以优化它们标记天然产物、天然产物前体和来自总合成努力的中间体的集合的能力（方案1）。酰胺键的形成可能是最直观的缀合手段，[16]因为原核和真核细胞中的代谢过程已被证明在体内进行相当的酰胺化。[17-20]对含羧酸的天然产物冈田酸[21-22]和（S）-疏螺旋体素[23-25]进行了研究。
Nearly forty years after Wieland’s studies on the labeling of phalloidin with fluorescein isothiocyanate,[1] fluorescent modifications have become an important tool in the study of natural product biology.[2–5] Since Wieland’s efforts, there has been a profound increase in the number of labels. This increase has led to the creation of a complex compendium of natural product probes, as illustrated by the generation of Aa–Dd from the appendage of labels A–D to natural products a–d (Figure 1 A). The resulting compendium Aa–Dd now contains modifications within three distinct variables: the label, linkage, and natural product. While variations in the label and linker might be of use in activity optimization, the extended diversity in Aa–Dd profoundly inhibits the ability to conduct comparative analyses. By using a single label and a common set of linkers, the number of variables displayed within a collection of natural product probes is reduced. As a result, data collected from the refined set of probes Da–Dd are homogenized (Figure 1B). Our studies began by identifying a route for labeling a diverse set of natural products from a single dye. In choosing the optimal dye, we considered a combination of criteria such as good photophysical properties, sufficient cellular uptake, lack of subcellular localization and lack of intrinsic biological activity. For small-molecule conjugation, carbocyanine dyes and rhodamine dyes, including rhodamine 123 and rhodamine B, are taken up into the mitochondria, while other common labels, such as cyanine dyes or BODIPY dyes localize in the nucleus and endoplasmic reticulum (ER), respectively. Ethyl 7-dimethylaminocoumarin-4-acetate (1)[6–7] was chosen as the central dye for these studies based on its lack of biological activity. Coumarin 1 is readily taken up into cells, however, it does not localize and can be removed by washing the cells. Additionally, its water-solubility,[8] small size,[9] availability,[10] photostability,[11] and photophysical properties (lex= 370 nm, e= 22 000M À1cmÀ1, lem= 459 nm, F= 0.1–0.4) were favorable to the development of synthetic applications, while providing viable fluorescence for a diverse set of photophysical experiments.[12] Coumarin 1 was converted into a set of functionalized labels 2–9 by using procedures developed by members of this team.[7–8] By using this route, 10–100 g batches of 1 were routinely prepared from m-(N, N)-dimethylaminophenol and diethyl 1, 3-acetonedicarboxylate and converted to acid 2. A set of labels was prepared from either ester 1 or acid 2. As depicted in Scheme 1, the coupling of 2 with glycine tert-butyl ester, followed by deprotection with TFA and esterification with N-hydroxysuccinimide provided ester 3 in 25% overall yield. Comparable amide couplings were also effective in synthesizing labels 4–6 and 8–9 (Scheme 1). Alternatively, labels 3–9 could be prepared by thiol-assisted addition of a primary amine to ester 1, as illustrated by the reaction of cysteamine with 1 to afford 7.[13–14] The addition of 5–20 mol% of sodium thiophenoxide, sodium ethanethiolate, or groupIV metal alkoxides [15] can be used to facilitate this process. With labels 2–9 in hand, our efforts turned to screening reaction conditions so as to optimize their ability to tag a collection of natural products, natural product precursors, and intermediates from total synthetic efforts (Scheme 1). Amide-bond formation was perhaps the most intuitive means of conjugation,[16] as metabolic processes in prokaryotic and eukaryotic cells have been shown to conduct comparable amidation in vivo.[17–20] The carboxylic acid-containing natural products, okadaic acid [21–22] and (À)-borrelidin [23–25] were …