2’,3’-Protected Nucleotides as Building Blocks for Enzymatic de novo RNA Synthesis
2',3'-保护核苷酸作为酶促从头 RNA 合成的构建模块
首次发布：2024 年 2 月 25 日 https://doi.org/10.1002/chem.202400137 引用次数：1
Graphical Abstract 图形摘要
RNA one by one. Here, we present a first step towards controlled enzymatic RNA synthesis. We have explored the possibility of a simple protection step of the vicinal cis-diol moiety to temporarily block ribonucleotides. We demonstrate that pyrimidine nucleotides protected with acetals, particularly 2′,3′-O-isopropylidene, are well-tolerated by the template-independent RNA polymerase PUP and highly efficient coupling reactions can be achieved within minutes.
RNA 逐一合成。在此，我们展示了迈向可控酶促 RNA 合成的第一步。我们探索了一种简单的邻位顺式二醇基团保护步骤，以暂时阻断核糖核苷酸的可能性。我们证明，用缩醛（特别是 2′,3′-O-异丙叉）保护的嘧啶核苷酸被模板非依赖性 RNA 聚合酶 PUP 良好耐受，并且可以在几分钟内实现高效的偶联反应。
Abstract 摘要
Besides being a key player in numerous fundamental biological processes, RNA also represents a versatile platform for the creation of therapeutic agents and efficient vaccines. The production of RNA oligonucleotides, especially those decorated with chemical modifications, cannot meet the exponential demand. Due to the inherent limits of solid-phase synthesis and in vitro transcription, alternative, biocatalytic approaches are in dire need to facilitate the production of RNA oligonucleotides. Here, we present a first step towards the controlled enzymatic synthesis of RNA oligonucleotides. We have explored the possibility of a simple protection step of the vicinal cis-diol moiety to temporarily block ribonucleotides. We demonstrate that pyrimidine nucleotides protected with acetals, particularly 2′,3′-O-isopropylidene, are well-tolerated by the template-independent RNA polymerase PUP (polyU polymerase) and highly efficient coupling reactions can be achieved within minutes – an important feature for the development of enzymatic de novo synthesis protocols. Even though purines are not equally well-tolerated, these findings clearly demonstrate the possibility of using cis-diol-protected ribonucleotides combined with template-independent polymerases for the stepwise construction of RNA oligonucleotides.
除了在众多基础生物过程中扮演关键角色外，RNA 还为治疗剂和高效疫苗的研发提供了一个多功能平台。RNA 寡核苷酸，尤其是那些经过化学修饰的，其生产无法满足指数级增长的需求。鉴于固相合成和体外转录的固有局限性，迫切需要替代的生物催化方法来促进 RNA 寡核苷酸的生产。在此，我们向可控的酶促合成 RNA 寡核苷酸迈出了第一步。我们探索了通过简单的保护步骤暂时封闭核糖核苷酸中相邻顺式二醇基团的可能性。我们证明，用缩醛（特别是 2′,3′-O-异丙叉）保护的嘧啶核苷酸能被模板非依赖性 RNA 聚合酶 PUP（聚 U 聚合酶）良好耐受，并且能在几分钟内实现高效的偶联反应——这是开发酶促从头合成方案的一个重要特性。 尽管嘌呤的耐受性不尽相同，这些发现清楚地展示了利用顺式二醇保护的核糖核苷酸与非模板依赖型聚合酶相结合，逐步构建 RNA 寡核苷酸的可能性。
1 Introduction 1 引言
In the early 1990s, RNA oligonucleotides were conceived as promising candidates for the development of therapeutics yet difficult to handle, poorly understood, and challenging to synthesize. The recent FDA-approval of several siRNA drugs and the advent of mRNA vaccines (rewarded by the Nobel Prize in Physiology and Medicine in 2023) have dramatically changed this early vision and thrust RNA in the forefront of drug discovery.1 Nonetheless, RNA therapeutics could benefit from a better understanding of the mechanisms underlying existing modalities such as siRNAs, anti-miR, or mRNA vaccines but also from poorly understood RNA molecules such as non-coding RNAs2 and other regulatory RNAs3 as well as the effect of each of the >100 naturally occurring RNA modifications.4 A better understanding of RNA biology however, is tightly tied to our capacity at producing RNA oligonucleotides with or without chemical modifications in both high yields and purity. So far, short RNAs are mainly produced chemically by assembling phosphoramidite building blocks on immobilized nucleosides.5 While this method represents the workhorse for the production of therapeutic siRNAs and antisense oligonucleotides,6 developing more efficient and sustainable alternatives granting access to longer RNA oligonucleotides remains an important challenge.7 Indeed, success of solid-phase assembly of RNA oligonucleotides relies on the careful choice of the 2’-O-protecting groups which directly affects the reactivity of phosphoramidite building blocks.8 In addition, chemical synthesis is highly efficient for shorter (i. e. 10–60 nt) sequences, but yields of production rapidly decrease with increasing length of the oligonucleotides.9 Chemical synthesis of natural and modified RNA is also impinged by sustainability issues since poor atom economy and the necessity for intricate protecting group strategies generates large amounts of chemical waste.10 Alternatively, RNAs can be produced by in vitro transcription reactions using nucleoside triphosphates (NTPs) and RNA polymerases.11 This highly potent method can be used for the identification of functional nucleic acids via the SELEX (Systematic Evolution of Ligands by EXponential enrichment) protocol,12 but also for the industrial-scale production of therapeutic oligonucleotides as showcased by mRNA vaccines during the COVID 19 pandemics.13 Notwithstanding these favorable assets, enzymatic synthesis depends on the recognition of modified nucleotides by polymerases and control of the localization of chemically altered nucleotides within the sequence is limited. Hence, alternative methods for the production of modified and unmodified RNA oligonucleotides are in dire need.7 Biocatalysis represents an alluring option to improve the sustainability, the purity and yield of oligonucleotide production while simultaneously enabling the control of modification localization. Various biocatalytic strategies have been recently proposed for nucleic acid de novo synthesis.14 In controlled enzymatic synthesis, temporarily blocked nucleotides are added sequentially at the 3’-termini of immobilized primers by (mainly template-independent) polymerases. Robust protocols have been devised for the de novo synthesis of DNA by controlled enzymatic synthesis,9b, 15 but this approach has been vastly overlooked for the production of chemically modified oligonucleotides16 as well as RNAs.17 Challenges in the development of versatile and reliable protocols for efficient controlled enzymatic synthesis of RNA oligonucleotides reside in the identification of suitable sugar and/or nucleobase protecting groups which need to be compatible with RNA polymerases as well as sufficiently robust to withstand enzymatic synthesis but at the same time labile enough to allow for facile and rapid deprotection. Herein, we have explored the possibility of using protecting groups for the cis-2’,3’-diol of ribonucleosides to generate temporarily blocked nucleotides compatible with controlled enzymatic RNA synthesis. Of the different protecting group strategies that were evaluated, we found that 2′,3′-O-isopropylidene-blocked pyrimidine triphosphates were incorporated quantitatively within minutes by the template-independent RNA polymerase polyuridine polymerase (PUP). The corresponding purine nucleotides were more reluctant at acting as substrates for both the PUP and polyadenosine polymerase (PAP) but could still be useful synthons for the production of RNA under controlled enzymatic synthesis conditions. Collectively, we demonstrate a first step towards the step-by-step production of RNA oligonucleotides by incorporating cis-2’,3’-diol-protected nucleotides into RNA using template-independent polymerases. Cis-diol reversible terminators could be directly amenable to the production of unmodified, as well as base- and phosphate-modified RNA oligonucleotides.
20 世纪 90 年代初，RNA 寡核苷酸被视为极具潜力的治疗药物开发候选者，但当时它们难以操控、理解不足且合成极具挑战性。近期，FDA 批准了数款 siRNA 药物以及 mRNA 疫苗的出现（这一成就荣获 2023 年诺贝尔生理学或医学奖），彻底颠覆了早期的看法，将 RNA 推向了药物发现的最前沿。 1 尽管如此，RNA 疗法仍可从更深入理解现有模式如 siRNA、抗 miR 或 mRNA 疫苗的机制中获益，同时，对非编码 RNA 2 及其他调控 RNA 3 等知之甚少的 RNA 分子，以及超过 100 种天然存在的 RNA 修饰各自的作用 4 ，也需进一步探索。然而，对 RNA 生物学的深入理解，紧密关联于我们能否高效、高纯度地生产带有或不带化学修饰的 RNA 寡核苷酸。迄今为止，短链 RNA 主要通过化学方法，在固定化的核苷上组装亚磷酰胺单体来生产。 5 虽然此方法是生产治疗性 siRNA 和反义寡核苷酸的主力， 6 但开发更高效、可持续的替代方案以获得更长的 RNA 寡核苷酸仍是一个重要挑战。 7 实际上，RNA 寡核苷酸的固相组装成功与否，关键在于 2'-O-保护基团的精心选择，这直接影响着磷酰胺单体构建模块的反应活性。 8 此外，化学合成对于较短（即 10-60 个核苷酸）的序列极为高效，但随着寡核苷酸长度的增加，产量迅速下降。 9 天然及修饰 RNA 的化学合成还受到可持续性问题的制约，因为原子经济性差及需要复杂的保护基团策略，导致大量化学废物的产生。 10 另一种方法是，可以通过使用核苷三磷酸（NTPs）和 RNA 聚合酶进行体外转录反应来生产 RNA。 11 这种高效的方法可用于通过 SELEX（指数富集配体系统进化）协议识别功能性核酸， 12 也可用于工业规模生产治疗性寡核苷酸，如 COVID-19 疫情期间 mRNA 疫苗所展示的那样。 13 尽管有这些有利条件，酶促合成依赖于聚合酶对修饰核苷酸的识别，且对序列中化学修饰核苷酸位置的控制有限。因此，迫切需要生产修饰和未修饰 RNA 寡核苷酸的替代方法。 7 生物催化为提高寡核苷酸生产的可持续性、纯度和产量提供了一个诱人的选择，同时还能控制修饰的位置。最近，已经提出了多种生物催化策略用于核酸的从头合成。 14 在受控的酶促合成中，暂时封闭的核苷酸通过（主要是模板非依赖性的）聚合酶依次添加到固定引物的 3'末端。 已经设计了稳健的协议，通过受控的酶促合成从头合成 DNA， 9b, 15 但这种方法在化学修饰的寡核苷酸 16 以及 RNA 的生产中被极大地忽视了。 17 开发通用且可靠的协议以高效进行 RNA 寡核苷酸受控酶促合成的挑战在于识别合适的糖和/或核碱基保护基团，这些基团需要与 RNA 聚合酶兼容，并且足够坚固以承受酶促合成，同时又要足够不稳定以允许快速和容易的去保护。在此，我们探索了使用保护基团来暂时封闭核糖核苷的顺式 2',3'-二醇的可能性，以生成与受控酶促 RNA 合成兼容的核苷酸。在评估的不同保护基团策略中，我们发现 2',3'-O-异丙叉封闭的嘧啶三磷酸盐在几分钟内被模板非依赖性 RNA 聚合酶聚尿苷聚合酶（PUP）定量地掺入。 相应的嘌呤核苷酸作为 PUP 和多腺苷酸聚合酶（PAP）的底物时表现较为勉强，但在受控的酶促合成条件下，它们仍可作为生产 RNA 的有用合成子。总的来说，我们展示了通过使用模板非依赖性聚合酶将顺式-2',3'-二醇保护的核苷酸整合到 RNA 中，逐步生产 RNA 寡核苷酸的第一步。顺式二醇可逆终止剂可直接用于生产未修饰的以及碱基和磷酸修饰的 RNA 寡核苷酸。
2 Results 2 结果
2.1 Design and Synthesis of Blocked Nucleosides and Nucleotides
2.1 阻断核苷和核苷酸的设计与合成
The design of RNA nucleotides equipped with reversible blocking groups presents an additional challenge compared to the corresponding DNA or xenonucleic acids (XNA) nucleoside triphosphates. Indeed, strictly speaking only the 3’-OH moiety of the nucleotide requires a masking group to prevent further incorporation events from occurring but this in turn requires a selective protection of the vicinal cis-diol pattern which is notoriously challenging.8a, 18 This inherent difficulty mainly stems from the ease of 2′,3′-migration during synthesis, lower reactivity of the secondary alcohols at the 2’/3’-positions compared to 5’-OH, and the need for orthogonality with other blocking groups on the nucleosidic scaffold.19 In order to reduce synthetic efforts and streamline the production of blocked RNA nucleotides, we opted for a single, cis-2’,3’-diol-protection step followed by conversion to the corresponding nucleotides (Scheme 1). We also surmised that dual protection of hydroxyl moieties would prevent the formation of 2’-5’-branched RNA from occurring due to the transient blocking of both reactive moieties.
与相应的 DNA 或异源核酸（XNA）核苷三磷酸相比，配备可逆阻断基团的 RNA 核苷酸设计提出了额外的挑战。确实，严格来说，只有核苷酸的 3'-OH 部分需要掩蔽基团以防止进一步的掺入事件发生，但这反过来又需要对邻近的顺式二醇模式进行选择性保护，这众所周知是具有挑战性的。 8a, 18 这种固有的困难主要源于合成过程中 2',3'-迁移的容易性，2'/3'-位次级醇的反应性低于 5'-OH，以及需要与核苷支架上的其他阻断基团正交。 19 为了减少合成工作量并简化阻断 RNA 核苷酸的生产，我们选择了单一的顺式 2',3'-二醇保护步骤，然后转化为相应的核苷酸（方案 1）。我们还推测，羟基的双重保护将防止由于两个反应性基团的暂时阻断而形成 2'-5'-分支 RNA。
Description of the rationale involving cis-2’,3’-diol-protection of ribonuclosides/ribonucleotides and chemical structures of the explored protecting groups.
核糖核苷/核糖核苷酸顺式-2',3'-二醇保护原理的描述及所探索保护基团的化学结构。
In addition to opting for a cis-2’,3’-diol-protection strategy, we decided to explore various typical, well-established blocking groups commonly used in RNA chemistry such as boronic acids,20 esters,21 and acetals (Scheme 1).22
除了选择顺式-2',3'-二醇保护策略外，我们还决定探索 RNA 化学中常用的各种典型且成熟的阻断基团，如硼酸、 20 酯、 21 和缩醛（方案 1）。 22
First, we turned our attention to the formation of uridine nucleosides functionalized with reversible boronic esters (Scheme 1). Boronic acids selectively react with vicinal diols to form five-membered cyclic esters, which made them popular transient protecting groups in synthetic routes to modified ribonucleosides.20a, 20d Consequently, boronic esters appears as interesting alternatives for enzymatic RNA synthesis due to their compatibility with vicinal diols, relative stability in aqueous conditions, and ease of deprotection. The presence of bulky groups or substituted aromatic residues can stabilize boronic esters against hydrolytic degradation.23 As a compromise between enhanced stability and compatibility with polymerase-mediated synthesis, we dwelled on aryl boronates as a choice for potential temporary masking groups. In addition, we considered various substitution patterns on the aromatic moiety of phenyl boronic acid (1 a) including electron withdrawing elements (1 d) and extended aromatic systems such as naphtyl (1 b), and pyrenyl (1 c) moieties (Scheme 2). The synthesis of the corresponding boronic esters was straightforward and involved treatment of uridine with the corresponding boronic acid in the presence of MgSO4 to quench the released water molecule. After precipitation in acetonitrile, the corresponding esters were isolated in moderate (~40 %) to good (~80 %) yields. Boronic esters often display limited stability in aqueous media due to hydrolytic degradation via transesterification or protodeborination.20a, 24 Hence, we carried out an 1H NMR comparative study to investigate whether nucleosides 1 a–1 d are stable in water, an important prerequisite for storage and enzymatic reactions with the corresponding nucleotides. To do so, nucleosides 1 a–1 c were incubated in a DMSO-d6:D2O mixture at room temperature for 90 min and 1H NMR spectra were compared to that of unreacted uridine and the parent boronic esters (Figure S1–S3, Supporting Information). Under these conditions, all nucleosides were cleanly and completely converted to uridine. Hence, while boronic esters are clearly not suitable as temporary masking groups in enzymatic RNA synthesis, they might represent versatile, transient protecting groups for the preparation of RNA nucleoside analogues given the ease of preparation and the mild deprotection conditions.
首先，我们将注意力转向了带有可逆硼酸酯功能化的尿苷核苷的形成（方案 1）。硼酸选择性地与邻二醇反应生成五元环状酯，这使其成为合成修饰核糖核苷酸路线中流行的临时保护基团。 20a, 20d 因此，硼酸酯因其与邻二醇的兼容性、在水性条件下的相对稳定性以及易于脱保护，成为酶促 RNA 合成中引人注目的替代品。大体积基团或取代芳香残基的存在可以稳定硼酸酯，防止水解降解。 23 在增强稳定性和与聚合酶介导合成兼容性之间权衡后，我们选择了芳基硼酸酯作为潜在的临时掩蔽基团。此外，我们还考虑了苯硼酸（1a）芳香部分上的各种取代模式，包括吸电子元素（1d）以及扩展的芳香体系，如萘基（1b）和芘基（1c）部分（方案 2）。 相应硼酸酯的合成过程直接明了，涉及在 MgSO₄存在下用相应的硼酸处理尿苷以淬灭释放的水分子。在乙腈中沉淀后，相应的酯以中等（约 40%）至良好（约 80%）的收率被分离出来。由于通过酯交换或脱硼化反应导致的水解降解，硼酸酯在水介质中通常表现出有限的稳定性。因此，我们进行了¹H NMR 对比研究，以探讨核苷 1a-1d 在水中是否稳定，这是储存和与相应核苷酸进行酶促反应的重要前提。为此，将核苷 1a-1c 在室温下于 DMSO-d₆:D₂O 混合物中孵育 90 分钟，并将¹H NMR 谱图与未反应的尿苷和母体硼酸酯的谱图进行比较（图 S1-S3，支持信息）。在这些条件下，所有核苷均干净且完全地转化为尿苷。 因此，尽管硼酸酯显然不适合作为酶促 RNA 合成中的临时掩蔽基团，但由于其易于制备和温和的去保护条件，它们可能代表了制备 RNA 核苷类似物的多功能、瞬态保护基团。
Synthesis of uridine nucleosides with boronic ester protecting groups. Reagents and conditions: i) a) RB(OH)2, pyridine, reflux, 4–5 h, MgSO4; b) Precipitation in CH3CN; 39 % (1 a), 43 % (1 b), 87 % (1 c), 74 % (1 d).
带有硼酸酯保护基团的尿苷核苷的合成。试剂与条件：i) a) RB(OH) 2 ，吡啶，回流，4-5 小时，MgSO 4 ；b) 在 CH 3 CN 中沉淀；39% (1a)，43% (1b)，87% (1c)，74% (1d)。
We next considered the possibility of using esters as temporary blocking groups for RNA nucleotides. We have previously reported the compatibility of locked nucleic acid (LNA) nucleotides equipped with various 3’-O-ester moieties.16b, 16c From these studies, we concluded that 1) esters such as allyl or nitrobenzoyl were not sufficiently robust to be considered as blocking groups of nucleotides; 2) very bulky groups such as mesitoyl were not tolerated by DNA polymerases; and 3) esters of intermediate size and robustness such as benzoyl and pivaloyl were compatible with enzymatic synthesis despite some hydrolysis caused by the esterase activity of DNA polymerases.16a, 25 Based on these considerations, we opted to equip uridine with benzoyl (2 b) and pivaloyl (2 c) esters as temporary blocking groups. In addition, since esterase activity has been observed for DNA but not for RNA polymerases, we also considered acetyl (2 a) as a blocking group. We therefore first prepared the suitably protected nucleosides 2 a–2 c by treating directly commercially available 5’-O-DMTr-uridine with either acyl chlorides or acetic anhydride under optimized conditions to minimize N3-acylation (Scheme 3). The resulting nucleosides were then treated under mild acidic conditions to remove the trityl protecting group and the resulting nucleosides 3 a–3 c were converted to triphosphates using either the method based on chlorophosphorinone26 (for 4 a and 4 b) or POCl327 (for 4 c) as phosphorylation reagents. Interestingly, N3-benzoyl-bis-2’,3’-O-benzoyl-uridine, which arises as a side-product of the esterification reaction, was highly refractory to the phosphorylation reactions and no conversion to triphosphate could be observed.
接下来，我们探讨了使用酯类作为 RNA 核苷酸临时封闭基团的可能性。我们先前已报道了配备各种 3'-O-酯基团的锁定核酸（LNA）核苷酸的兼容性。 16b, 16c 通过这些研究，我们得出以下结论：1）如烯丙基或硝基苯甲酰等酯类作为核苷酸的封闭基团不够稳定；2）DNA 聚合酶无法容忍如 mesitoyl 等非常庞大的基团；3）中等大小和稳定性的酯类，如苯甲酰基和新戊酰基，尽管 DNA 聚合酶的酯酶活性会导致一定程度的水解，但仍与酶促合成兼容。 16a, 25 基于这些考虑，我们选择将尿苷配备苯甲酰基（2b）和新戊酰基（2c）酯作为临时封闭基团。此外，由于已观察到 DNA 而非 RNA 聚合酶具有酯酶活性，我们也考虑了乙酰基（2a）作为封闭基团。因此，我们首先通过直接处理市售的 5'-O-DMTr-尿苷，在优化条件下使用酰氯或乙酸酐进行反应，以最小化 N3-酰化，制备了适当保护的核苷 2a–2c（方案 3）。 所得核苷随后在温和酸性条件下处理以去除三苯甲基保护基团，并将所得核苷 3a–3c 通过基于氯磷酰酮 26 （用于 4a 和 4b）或 POCl 3 27 （用于 4c）作为磷酸化试剂的方法转化为三磷酸酯。有趣的是，作为酯化反应副产物产生的 N3-苯甲酰基-双-2',3'-O-苯甲酰基-尿苷对磷酸化反应表现出极高的抗性，未能观察到其转化为三磷酸酯。
Synthesis of ester-blocked nucleosides (3 a–3 c) and nucleotides (4 a–4 c). Reagents and conditions: i) Ac2O, pyridine, rt, 12 h, 85 % (2 a), ii) BzCl, DMAP, DCM, Et3N, rt, 5.5 h, 82 % (2 b), iii) PivCl, pyridine, rt to 50 °C, 7 days, 40 % (2 c); iv) TFA, DCM, rt, 30 min, 80 % (3 a) and 56 % (3 c), v) DCA (2 %) in DCM, rt, 25 min, quant. (3 b); vi) 1) 2-chloro-1,3,2-benzodioxaphosphorin-4-one (1.2 eq.), Pyridine/dioxane (2 : 1), 0 °C, 4.5 h; 2) (nBu3NH)2H2P2O7 (1.2 eq.), nBu3N (1.2 eq.), DMF, 0 °C, 2.5 h; 3) I2 (1.2 eq.), pyridine/H2O (95 : 5), rt, 20 min; 1 % (4 a) and 22 % (4 b), vii) 1) POCl3 (1.1 eq.), PO(OMe)3 , 0 °C, 4 h, 2) (nBu3NH)2H2P2O7 (1.1 eq.), nBu3N (1.1 eq.), DMF, 0 °C, 1 h; 3) TEAB 1 M, rt, 30 min, 4 % (4 c).
酯封闭核苷（3a–3c）和核苷酸（4a–4c）的合成。试剂及条件：i) 乙酸酐，吡啶，室温，12 小时，85%（2a）；ii) 苯甲酰氯，DMAP，二氯甲烷，三乙胺，室温，5.5 小时，82%（2b）；iii) 新戊酰氯，吡啶，室温至 50°C，7 天，40%（2c）；iv) 三氟乙酸，二氯甲烷，室温，30 分钟，80%（3a）和 56%（3c）；v) 2%二氯乙酸于二氯甲烷中，室温，25 分钟，定量（3b）；vi) 1) 2-氯-1,3,2-苯并二氧磷杂环戊烷-4-酮（1.2 当量），吡啶/二氧六环（2:1），0°C，4.5 小时；2) 正丁胺磷酸盐（1.2 当量），正丁胺（1.2 当量），DMF，0°C，2.5 小时；3) 碘（1.2 当量），吡啶/水（95:5），室温，20 分钟；1%（4a）和 22%（4b）；vii) 1) 三氯氧磷（1.1 当量），磷酸三甲酯，0°C，4 小时；2) 正丁胺磷酸盐（1.1 当量），正丁胺（1.1 当量），DMF，0°C，1 小时；3) 1M TEAB，室温，30 分钟，4%（4c）。
Lastly, besides esters, acetals such as isopropylidene are common protecting groups for the vicinal cis-diol pattern of RNA nucleosides and nucleotides. These groups can easily be installed on RNA nucleosides by reacting the corresponding ketones and diols under acidic conditions or by transketalization.28 In addition to facile preparation, we surmised that acetals protecting groups might resist against a potential esterase activity of RNA polymerases. Consequently, we set out to synthesize uridine nucleosides and nucleotides equipped with 2′,3′-O-isopropylidene (5 a and 6 a), 2′,3′-O-cyclohexylidene (5 b and 6 b), and 2′,3′-O-benzylidene (5 c and 6 c) moieties (Scheme 4). We also included a 2-(methoxycarbonyl)ethylidene (or Moc-ethylidene) acetal protecting group (10 and 6 d) since this moiety can be cleaved under basic rather than (often strongly) acidic conditions typical for ketal removal (Scheme 5).29 We also included commercially available 2’,3’-O-trinitrophenyl uridine 5’-triphosphate (TNP-UTP) in this study. While TNP-nucleotides are usually employed as probes to target nucleotide-binding proteins such as enzymes, receptors and structural proteins with an affinity in the micromolar range,30 they have never been considered as substrates for template-independent RNA polymerases and strictly speaking are also acetal-blocked nucleotides.
最后，除了酯类，如异亚丙基等缩醛也是 RNA 核苷和核苷酸中邻位顺式二醇模式的常见保护基团。这些基团可以通过在酸性条件下使相应的酮和二醇反应或通过转缩醛化反应轻松地安装在 RNA 核苷上。 28 除了易于制备外，我们推测缩醛保护基团可能抵抗 RNA 聚合酶潜在的酯酶活性。因此，我们着手合成了带有 2′,3′-O-异亚丙基（5a 和 6a）、2′,3′-O-环己亚基（5b 和 6b）和 2′,3′-O-苯亚甲基（5c 和 6c）部分的尿苷核苷和核苷酸（方案 4）。我们还引入了 2-(甲氧羰基)乙亚基（或 Moc-乙亚基）缩醛保护基团（10 和 6d），因为该部分可以在碱性而非（通常强）酸性条件下被裂解，这是典型的缩酮去除条件（方案 5）。 29 我们还将市售的 2’,3’-O-三硝基苯基尿苷 5’-三磷酸（TNP-UTP）纳入了本研究。 虽然 TNP-核苷酸通常被用作探针，以靶向微摩尔亲和力范围内的核苷酸结合蛋白，如酶、受体和结构蛋白， 30 但它们从未被视为模板非依赖性 RNA 聚合酶的底物，严格来说，它们也是缩醛封闭的核苷酸。
Synthesis of acetal-blocked nucleosides (5 a–5 c) and nucleotides (6 a–6 c). Reagents and conditions: i) 2,2-dimethoxypropane, APTS, acetone, 0 °C to reflux, 2 h, 72 % (5 a); ii) cyclohexanone, APTS, 50 °C, 4 h, 82 % (5 b); iii) benzaldehyde, APTS, molecular sieves, 70 °C, 12 h, 37 % (5 c); iv) 1) 2-chloro-1,3,2-benzodioxaphosphorin-4-one (1.2 eq.), Pyridine/dioxane (2 : 1), 0 °C, 4 h30; 2) (nBu3NH)2H2P2O7 (1.2 eq.), nBu3N (1.2 eq.), DMF, 0 °C, 2 h30; 3) I2 (1.2 eq.), pyridine/H2O (95 : 5), rt, 20 min; 6 % (6 a) and 5 % (6 b), v) 1) POCl3 (1.1 eq.), PO(OMe)3 , 0 °C, 4 h, 2) (nBu3NH)2H2P2O7 (1.1 eq.), nBu3N (1.1 eq.), DMF, 0 °C, 1 h; 3) TEAB 1 M, rt, 30 min, 16 % (6 a), 5 % (6 b), and 6 % (6 c).
合成缩醛保护的核苷（5a–5c）和核苷酸（6a–6c）。试剂及条件：i) 2,2-二甲氧基丙烷，APTS，丙酮，0°C 至回流，2 小时，72%（5a）；ii) 环己酮，APTS，50°C，4 小时，82%（5b）；iii) 苯甲醛，APTS，分子筛，70°C，12 小时，37%（5c）；iv) 1) 2-氯-1,3,2-苯并二氧磷杂环戊烷-4-酮（1.2 当量），吡啶/二氧六环（2:1），0°C，4 小时 30 分钟；2) (nBu 3 NH) 2 H 2 P 2 O 7 （1.2 当量），nBu 3 N（1.2 当量），DMF，0°C，2 小时 30 分钟；3) I 2 （1.2 当量），吡啶/H 2 O（95:5），室温，20 分钟；6%（6a）和 5%（6b），v) 1) POCl 3 （1.1 当量），PO(OMe) 3 ，0°C，4 小时，2) (nBu 3 NH) 2 H 2 P 2 O 7 （1.1 当量），nBu 3 N（1.1 当量），DMF，0°C，1 小时；3) TEAB 1 M，室温，30 分钟，16%（6a），5%（6b）和 6%（6c）。
Synthesis of acetal-blocked nucleoside 10 and nucleotide 6 d. Reagents and conditions: i) Boc2O, Et3N, DMAP, Pyridine, rt, 1 h, 88 % (7); ii) NH3 in MeOH (4 M), 5 °C to rt, 3 h, 86 % (8); iii) Methyl propynoate, DMAP, CH3CN, rt, 30 min, 78 % (9); iv) 5 % TFA in DCM, Et3SiH, 4–5 h, rt, 76 % (10); v) 1) 2-chloro-1,3,2-benzodioxaphosphorin-4-one (1.2 eq.), Pyridine/dioxane (2 : 1), 0 °C, 4 h30; 2) (nBu3NH)2H2P2O7 (1.2 eq.), nBu3N (1.2 eq.), DMF, 0 °C, 2 h30; 3) I2 (1.2 eq.), pyridine/H2O (95 : 5), rt, 20 min; 17 % (6 d).
缩醛保护的核苷 10 和核苷酸 6d 的合成。试剂及条件：i) Boc 2 O, Et 3 N, DMAP, 吡啶, 室温, 1 小时, 88% (7); ii) NH 3 在甲醇中(4 M), 5°C 至室温, 3 小时, 86% (8); iii) 丙炔酸甲酯, DMAP, CH 3 CN, 室温, 30 分钟, 78% (9); iv) 5% TFA 在 DCM 中, Et 3 SiH, 4-5 小时, 室温, 76% (10); v) 1) 2-氯-1,3,2-苯并二氧磷杂环戊烷-4-酮(1.2 当量), 吡啶/二氧六环(2:1), 0°C, 4 小时 30 分钟; 2) (nBu 3 NH) 2 H 2 P 2 O 7 (1.2 当量), nBu 3 N(1.2 当量), DMF, 0°C, 2 小时 30 分钟; 3) I 2 (1.2 当量), 吡啶/H 2 O(95:5), 室温, 20 分钟; 17% (6d)。
Synthetic routes to acetal-blocked nucleotides 6 a–6 c follows standard protocols. Briefly, uridine was converted to the corresponding 2′,3′-O-isopropylidene-protected nucleoside by treatment with 2,2-dimethoxypropane under acidic conditions in good yields (72 %). Nucleotide 6 a was obtained by phosphorylation with a P(III) reagent (6 % yield) or POCl3 (16 %). Nucleoside 5 b was obtained in good yields (72 %) by reacting uridine directly with cyclohexanone under acidic conditions and triphosphorylation led to the isolation of nucleotide 6 b in low yields (5 % regardless of the method). Similarly, when uridine was treated with benzaldehyde under acidic conditions, acetal 5 c could be obtained in acceptable (37 %) yields. An increase in reaction time did not improve the yield of this conversion since degradation of product was observed (data not shown). Finally, nucleotide 6 c was obtained by triphosphorylation of precursor 5 c in low yields (6 %) as an inseparable mixture of stereoisomers.
合成路线至缩醛保护的核苷酸 6a–6c 遵循标准协议。简而言之，尿苷在酸性条件下与 2,2-二甲氧基丙烷反应，以良好收率（72%）转化为相应的 2′,3′-O-异丙叉基保护的核苷。核苷酸 6a 通过 P(III)试剂（6%收率）或 POCl 3 （16%收率）磷酸化获得。核苷 5b 通过尿苷在酸性条件下直接与环己酮反应以良好收率（72%）获得，而三磷酸化导致核苷酸 6b 以低收率（无论采用何种方法均为 5%）分离。类似地，当尿苷在酸性条件下与苯甲醛反应时，可以以可接受的收率（37%）获得缩醛 5c。增加反应时间并未提高此转化的收率，因为观察到产物降解（数据未显示）。最后，核苷酸 6c 通过前体 5c 的三磷酸化以低收率（6%）获得，作为不可分离的立体异构体混合物。
Synthesis of Moc-ethylidene-protected UTP 6 d required first protection of the N3-position of the nucleobase 2 a,31 to avoid undesired alkylation.29 To do so, we converted nucleoside 2 a (Scheme 3) to the corresponding N3-Boc-protected analog 7 under standard conditions (Scheme 5). Removal of the O-acetyl protecting group under basic conditions32 followed by treatment with methyl propynoate in the presence of catalytic amounts of DMAP yielded nucleoside 9 as an inseparable mixture of stereoisomers.29 Deprotection of the DMTr and Boc protecting groups produced 10 in good yields. Deprotection of 9 with 5 % DCA in DCM for 45 min in the absence of Et3SiH only led to removal of the DMTr group in 50 % yield (data not shown). Finally, nucleotide 6 d was obtained by application of the Ludwig-Eckstein protocol in moderate yields (17 %).
合成 Moc-乙叉保护的 UTP 6d 首先需要对核碱基 2a 的 N3 位进行保护， 31 以避免不希望的烷基化反应。 29 为此，我们在标准条件下（方案 5）将核苷 2a（方案 3）转化为相应的 N3-Boc 保护类似物 7。在碱性条件下 32 去除 O-乙酰保护基团后，接着在催化量的 DMAP 存在下与丙炔酸甲酯反应，得到核苷 9，其为不可分离的立体异构体混合物。 29 去除 DMTr 和 Boc 保护基团后，以良好收率得到 10。在无 Et 3 SiH 的情况下，用 5% DCA 在 DCM 中处理 9 45 分钟，仅以 50%的收率去除了 DMTr 基团（数据未显示）。最后，通过应用 Ludwig-Eckstein 方案，以中等收率（17%）获得了核苷酸 6d。
2.2 Biochemical Characterization of 2’,3’-O-Blocked UTPs
2.2 2',3'-O-阻断 UTP 的生化特性表征
With ester- and acetal protected uridine triphosphates 4 a–4 c and 6 a–6 d, respectively at hand, we next sought to evaluate their capacity at acting as substrates in the context of controlled RNA synthesis. To this effect, our group16c and the laboratory of Church17 have independently identified the template-independent PUP and PAP polymerases as suitable candidates. Indeed, these polymerases are capable of acting like the TdT does with 3’-O-blocked DNA nucleotides15a-15c, 33 and appear to be quite tolerant to modified nucleotides.34 Moreover, we surmised that PUP could be employed for the introduction of blocked pyrimidine nucleotides while PAP would be required for the incorporation of similarly modified purines. To verify this hypothesis, we first set out to compare the substrate tolerance of both RNA polymerases with canonical nucleotides. To do so, we performed primer extension (PEX) reactions using a 5’-FAM-labelled, 18 nucleotide long RNA primer (5’-FAM-CAG UCG GAU CGC AGU CAG-3’) and each individual rNTP along with each of the template-independent RNA polymerases (Figure S4A). Gel analysis (PAGE 20 %) reveals that PUP equally well-tolerates UTP and ATP as substrates since robust tailing activities could be observed albeit the reaction with ATP led to larger product dispersities and lower size averages, consistent with previous reports.35 On the other hand, the PUP was rather reluctant at accepting CTP and GTP as substrates, nonetheless several incorporation events could be observed with full consumption of the primer which would be sufficient for controlled enzymatic synthesis applications. Unsurprisingly, the PAP readily incorporates rA and rC nucleotides into RNA primers.36 Overall, the substrate preference for PUP decreases in the order rU~rA > rC > rG and rA > rC > rU > rG for PAP. We also investigated the effect of UTP concentration on the tailing reaction efficiency of the PUP (Figure S4B). As observed for the TdT,37 the tailing efficiency of the PUP-catalyzed reaction strongly increases with UTP concentration when enzyme and initiator (primer) are both kept at a constant concentration.
手头有了分别以酯和缩醛保护的尿苷三磷酸 4a–4c 和 6a–6d 后，我们接下来评估了它们在受控 RNA 合成中作为底物的能力。为此，我们团队 16c 和 Church 实验室 17 独立地识别出模板非依赖性的 PUP 和 PAP 聚合酶为合适候选者。确实，这些聚合酶能够像 TdT 处理 3'-O-封闭的 DNA 核苷酸 15a-15c, 33 那样工作，并且对修饰核苷酸表现出相当高的耐受性 34 。此外，我们推测 PUP 可用于引入封闭的嘧啶核苷酸，而 PAP 则需用于类似修饰的嘌呤核苷酸的掺入。为了验证这一假设，我们首先着手比较两种 RNA 聚合酶对标准核苷酸的底物耐受性。为此，我们使用 5'-FAM 标记的 18 核苷酸长 RNA 引物（5'-FAM-CAG UCG GAU CGC AGU CAG-3'）和各个 rNTP，以及每种模板非依赖性 RNA 聚合酶进行了引物延伸（PEX）反应（图 S4A）。 凝胶分析（20% PAGE）显示，PUP 对 UTP 和 ATP 作为底物表现出相似的耐受性，尽管与 ATP 的反应导致产物分散度更大且平均尺寸较小，这与之前的报告一致。 35 另一方面，PUP 在接受 CTP 和 GTP 作为底物时显得较为勉强，尽管如此，仍可观察到几次掺入事件，且引物被完全消耗，这对于受控的酶促合成应用已足够。不出所料，PAP 能轻松地将 rA 和 rC 核苷酸掺入 RNA 引物中。 36 总体而言，PUP 对底物的偏好按 rU~rA > rC > rG 的顺序递减，而 PAP 则为 rA > rC > rU > rG。我们还研究了 UTP 浓度对 PUP 尾部反应效率的影响（图 S4B）。正如在 TdT 中观察到的那样， 37 当酶和引发剂（引物）浓度保持恒定时，PUP 催化反应的尾部效率随 UTP 浓度的增加而显著提高。
We next turned to evaluate the substrate tolerance of the blocked nucleotides with template-independent RNA polymerases. First, we evaluated the possibility of using ester-protected UTPs 4 a-4 c as substrates for the PUP polymerase to mediate single incorporation events. Analysis of the products stemming from the reaction of 4 a with both PUP and PAP displayed a product distribution reminiscent of that obtained with unmodified UTP (Figure S5A). This result is consistent with the rather low hydrolytic stability observed for a related 3’-O-Ac-LNA−T nucleoside.16b Surprisingly, equipping UTP with benzoyl (Figure S6) and pivaloyl (Figure S5B) masking groups, which had been identified as suitable for single incorporation of LNA−T nucleotides,16b led to low conversion yields (<40 %) and production of n+2 and n+3 side-products or completely abrogated substrate acceptance by the polymerase, respectively.
接下来，我们转向评估带有模板非依赖性 RNA 聚合酶的封闭核苷酸的底物耐受性。首先，我们评估了使用酯保护的 UTPs 4a-4c 作为 PUP 聚合酶底物以介导单次掺入事件的可能性。对 4a 与 PUP 和 PAP 反应产物的分析显示，产物分布与未修饰 UTP 所得结果相似（图 S5A）。这一结果与观察到的相关 3'-O-Ac-LNA−T 核苷较低的水解稳定性一致。 16b 令人惊讶的是，给 UTP 装备上苯甲酰（图 S6）和新戊酰（图 S5B）掩蔽基团，这些基团已被确定为适合 LNA−T 核苷酸单次掺入， 16b 却导致了较低的转化率（<40%）及 n+2 和 n+3 副产物的生成，或完全抑制了聚合酶对底物的接受。
We next assayed the acetonide-modified nucleotides 6 a–6 e under similar PEX reactions with PUP and PAP. We first carried out PEX reactions by supplementing the mixtures with different divalent metal cofactors (Mn2+, Co2+, Mg2+, and Zn2+) and nucleotide 6 a. Gel analysis (Figure 1) revealed that full conversion of the primer to the corresponding n+1 product could be achieved with little formation of n+2 and degradation products when 6 b was incubated with PUP and Mn2+ for 5 h at 37 °C (Figure 1). Degradation might occur via an exonuclease activity of the PUP which has been suggested previously38 or due to pyrophosphorolysis triggered by the poor substrate tolerance of the polymerase for 6 b.39 This analysis also revealed that the reaction mixtures supplemented with manganese led to the highest conversion yields. We also confirmed the necessity of adding Mn2+ by performing PEX reactions without adding any metal cofactor (Figure S7A). Moreover, lowering the RNA primer concentration led to a large product distribution suggesting an increased rate of deprotection of the modified UTP (Figure S7B).
接下来，我们在类似的 PEX 反应中测试了丙酮化物修饰的核苷酸 6a-6e 与 PUP 和 PAP 的反应。首先，我们通过向混合物中添加不同的二价金属辅因子（Mn 2+ 、Co 2+ 、Mg 2+ 和 Zn 2+ ）以及核苷酸 6a 进行了 PEX 反应。凝胶分析（图 1）显示，当 6b 与 PUP 和 Mn 2+ 在 37°C 下孵育 5 小时后，引物可以完全转化为相应的 n+1 产物，几乎没有 n+2 和降解产物的形成（图 1）。降解可能是通过 PUP 的外切酶活性发生的，这在之前已被提出 38 ，或者是由于聚合酶对 6b 的底物耐受性差引发的焦磷酸解作用 39 。该分析还表明，添加锰的反应混合物转化率最高。我们还通过在不添加任何金属辅因子的情况下进行 PEX 反应，确认了添加 Mn 2+ 的必要性（图 S7A）。此外，降低 RNA 引物浓度导致产物分布广泛，表明修饰 UTP 的去保护速率增加（图 S7B）。
Gel (PAGE 20 %) analysis of PEX reaction of 2′,3′-O-cyclohexylidene-UTP 6 b. All reactions were incubated with 1 mM of 6 b, 20 pmol of RNA primer, 10 U of PUP, 1 mM of metal (Mn 2+, Co 2+, Mg 2+, Zn 2+ from left to right), 20 U RNase Murine Inhibitor varying reaction time (i. e. 1 h, 3 h, 5 h) at 37 °C. Control reactions were carried out in the presence of unmodified UTP for 1 h (T+) or primer only (P).
2′,3′-O-环己叉-UTP 6b 的 PEX 反应凝胶（PAGE 20%）分析。所有反应均与 1 mM 的 6b、20 pmol RNA 引物、10 U PUP、1 mM 金属离子（从左至右分别为 Mn 2+ 、Co 2+ 、Mg 2+ 、Zn 2+ ）、20 U RNase 鼠源抑制剂在不同反应时间（即 1 小时、3 小时、5 小时）于 37°C 下孵育。对照反应在未修饰 UTP 存在下进行 1 小时（T+）或仅含引物（P）。
We next evaluated the substrate tolerance of nucleotides 6 a and 6 c equipped with 2′,3′-O-isopropylidene and 2′,3′-O-benzylidene blocking groups, respectively under the best conditions identified for 6 b with the PUP and PAP polymerases (Figure 2). Reactions carried out with nucleotide 6 c and the PUP mainly led to the formation of the desired n+1 product with only little n+2 and n+3 by-products. However, conversion of the primer was only in the ~50–70 % range. On the other hand, PEX reactions with 6 a and catalysed by the PUP led to more complex product distributions even though the n+1 product also appeared as the main product. Expectedly, the PAP did not readily accept either of these nucleotides as substrates and produced the extended n+1 primer only in moderate yields (~40 %).
Gel (PAGE 20 %) analysis of PEX reaction of 2′,3′-O-isopropylidene-UTP 6 a and 2′,3′-O-benzylidene-UTP 6 c. All reactions were incubated with 1 mM of N*TP, 20 pmol of RNA primer, 10 U of PUP or PAP, 1 mM of Mn2+, 20 U RNase Murine Inhibitor varying reaction time (i. e. 1 h, 3 h, 5 h) at 37 °C. Control reactions were carried out in the presence of unmodified UTP for 1 h (T+) or primer only (P).
Encouraged by these initial results, we next sought to optimize the experimental conditions to exclusively produce primers extended by a single modified nucleotide and minimize the rate of hydrolysis of the blocking groups. To do so, we first considered the addition of the crowding agents DMSO and PEG to the reaction mixtures.40 When PEX reactions with 6 a–6 c were supplemented with either 10 % DMSO (Figure S8A), 20 % PEG, or a mixture of DMSO and PEG (Figure S8B), we observed a marked decrease in n+1 product formation for all conditions. After excluding crowding agents from the optimization parameters, we next evaluated the effect of nucleotide concentration on the outcome of the PEX reactions. We first varied the concentration of nucleotide 6 c in PEX reactions and observed that lower concentrations (i. e. 500 μM) led to near quantitative conversion of primer to the n+1 product (Fig S9). Even though this finding was surprising since the efficiency of tailing reactions with natural and modified nucleotides usually increases with nucleotide concentration,16b, 41 we lowered the range of concentration to 50–500 μM. Gratifyingly, conditions could be found where the primer was cleanly converted to the expected n+1 product for all three nucleotides 6 a–6 c (Figure 3). Interestingly, analysis of the PEX reactions conducted with 6 a revealed quantitative conversion to the extended primer within 60 min and at concentrations as low as 50 μM (Figure 3A). On the other hand, PEX reactions with nucleotide analogs 6 b and 6 c required longer reaction times and slightly higher concentrations to achieve similar conversion efficiencies.
Gel (PAGE 20 %) analysis of PEX reaction of N*TP-UTP 6 a, 6 b, 6 c and 6 d with concentration from 50 to 500 μM. All reactions were incubated with 10 pmol of RNA primer, 10 U PUP, 1 mM of Mn2+, 20 U RNase Murine Inhibitor varying reaction time A), B) and C) (i. e. 1 h, 3 h, 5 h) or D) (i. e. 30 min 1 h, 3 h) at 37 °C. Primer only (P).
We next applied similar reaction conditions to Moc-ethylidene-protected UTP 6 d (Figure 3D). Also for this nucleotide analogue, the highest primer conversion yields (~90 %) were achieved when the concentration of the incoming triphosphate was kept low (50 μM) albeit with longer reaction times (5 h). We also applied these low triphosphate conditions to ester-blocked nucleotides 4 a and 4 c, which were hydrolysed or not accepted by the polymerase, but to no avail since the PEX reactions either led to the formation of larger distribution of side-products or completely abrogated substrate acceptance by the polymerase, respectively (Figure S10). A similar outcome was observed when 2′,3′-O-trinitrophenyl-UTP 6 e was used in conjunction with the PUP suggesting a polymerase-mediated hydrolysis of the blocking group (Figure S11).
Having identified a protecting group that is compatible with PUP-mediated RNA synthesis, we next investigated the possibility of reducing the reaction time. Indeed, for efficient de novo RNA synthesis, suitable nucleotides need to be accepted by the polymerase, the reactions should be quantitative, and coupling times should be as short as possible. Hence, we carried out PEX reactions with nucleotide with reaction times in the 5 to 60 min range (Figure 4). This analysis revealed that 1) when nucleotide concentration was kept at 50 μM, the reaction was already complete within 15 min (Figure 4A) and 2) concentrations as low as 20 μM were still compatible with high yielding nucleotide incorporation, albeit with slightly longer reaction times (30 to 45 min; Figure 4B).
Gel (PAGE 20 %) analysis of PEX reaction of 2′,3′-O-isopropylidene UTP 6 a at lower concentration and for shorter reaction time. A) Concentration of 6 a from 50 to 200 μM at various reaction time (i. e. 15, 30, 45, 60 min), B) Concentration of 6 a at 10 or 20 μM at various reaction time (i. e. 5, 15, 30, 45, 60 min). All reactions were incubated with 10 pmol of RNA primer, 10 U PUP, 1 mM of Mn2+, 20 U RNase Murine Inhibitor at 37 °C. Primer only (P).
Overall, we have identified conditions that permit the production of n+1 elongated ssRNA primers with excellent yields using four different blocking groups, namely 2’,3’-O-isopropylidene, 2’,3’-O-cyclohexylidene, 2’,3’-O-benzylidene, and 2’,3’-O-moc-ethylidene. Of these, UTP analogue 6 a displayed the best compatibility with PUP-mediated synthesis since complete conversion of the primer could be achieved in less than 15 min of reaction, at low NTP concentration, and without the occurrence of side-products. Prompted by these encouraging results, we set out to synthesize the full set of 2’,3’-O-isopropylidene-blocked NTPs and evaluate the possibility of using these analogues in PUP and/or PAP-mediated RNA synthesis.
2.3 Synthesis and Biochemical Characterization of 2’,3’-O-Isopropylidene ATP, CTP, GTP, ITP
Since cytidine, guanosine, adenosine and inosine nucleosides equipped with a 2’,3’-O-isopropylidene group were all commercially available (11 a–d), we directly used these as substrates in the triphosphorylation reaction following Ludwig-Eckstein procedure to obtain the corresponding nucleotides 12 a–d in low to moderate yields (Scheme 6).
Synthesis of 2’,3’-O-isopropylidene CTP, GTP, ATP, ITP 12 a–d. 1) 2-chloro-1,3,2-benzodioxaphosphorin-4-one (1.2 eq.), Pyridine/dioxane (2 : 1), 0 °C, 4.5 h; 2) (nBu3NH)2H2P2O7 (1.2 eq.), nBu3N (1.2 eq.), DMF, 0 °C, 2.5 h; 3) I2 (1.2 eq.), pyridine/H2O (95 : 5), rt, 20 min: 1.5 % (12 a), 19 % (12 b), 1.5 % (12 c), 3 % (12 d).
With the complete set of 2’,3’-O-isopropylidene-blocked nucleotides 12 a–d at hand, we evaluated their substrate acceptance by the PUP and PAP polymerases. Cytidine nucleotide 12 a displayed a similar behavior to that of UTP 6 a (Figure 5), since n+1 product formation could be achieved in near quantitative yields within ~2 min of reaction with a relatively low NTP concentration (40 μM) and at room temperature (instead of 37 °C as for 6 a so as to suppress the formation of undesired n+2 product as much as possible). Nonetheless, the formation of very low amounts (<5%) of n+2 products was observed, suggesting a slightly reduced hydrolytic stability of the acetonide moiety when installed on CTP than UTP. Increasing both the reaction time and the nucleotide concentration were deleterious and led to the formation of complex product distributions (Figure S12 and S13).
Gel (PAGE 20 %) analysis of PEX reaction of 2′,3′-O-isopropylidene CTP 12 a. Concentration of 12 a at 30 or 40 μM at various reaction time (i. e. 1 min 45, 5, 15, 30 min). All reactions were incubated with 10 pmol of RNA primer, 10 U PUP, 1 mM of Mn2+, 20 U RNase Murine Inhibitor at 23 °C. Primer only (P).
Surprisingly, GTP 12 b was not readily accepted as a substrate by the PAP, irrespective of the concentration involved in the PEX reaction since only low (<10 %) conversion yields to the expected n+1 product could be achieved (Figure S14). Since the PUP displayed a certain capacity at incorporating unmodified rGTP into RNA (Figure S4A), we also investigated whether this enzyme could be coerced to produce the desired n+1 product. When GTP 12 b was supplied in low concentration (50 μM) to the PEX reaction mixtures, the primer was converted to the corresponding n+1 product in moderate yields (~50 %) and increasing the reaction time did not seem to improve the conversion rate (Figure S15). At higher concentrations, slower running products appear on the gel analysis of the reaction products suggesting a partial deblocking of the isopropylidene moiety. We also investigated whether other metal cofactors could positively influence the outcome of the PEX reactions with GTP 12 b, but to no avail (Figure S16). When ATP 12 c was evaluated as a substrate for the PAP polymerase, important product dispersities could be observed even at low nucleotide concentrations (Figure S17). A similar outcome was observed when the PAP was replaced by the PUP polymerase (Figure S18) and ATP 12 c with inosine analogue 12 d (Figure S19). The inherent instability of the 2’,3’-O-isopropylidene blocking group on purines is in stark contrast with that observed for pyrimidines where very little hydrolysis occurred. Intrigued by these observations, we sought to install a more robust blocking group on a purine nucleotide and evaluate whether hydrolysis could be prevented without impeding efficient incorporation. To do so, we prepared 2′,3′-O-cyclohexylidene-blocked ATP analogue 13 by application of the protocol outlined in Scheme 4 for uridine (see Experimental Section). With 13 at hand, we carried out PEX reactions with both the PAP and PUP polymerases (Figure S20 and S21). As expected, no hydrolysis of the more robust cyclohexylidene moiety was observed with both enzymes but n+1 product formation was very modest. Despite screening different experimental conditions, n+1 product formation by the PAP did not exceed 10–20 %, suggesting that bulkier blocking groups on purines are not well tolerated by these template-independent RNA polymerases.
2.4 Chemical Deprotection of Elongated Products
After biochemical verification of the compatibility of all blocked nucleotides with the PUP and PAP polymerases for the elongation of ssRNA primers, we carried out large-scale PEX reactions with nucleotides that were best tolerated. The resulting products were then analysed by MALDI-TOF after purification with an RNA Cleanup Kit. This analysis clearly demonstrated the chemical integrity of the n+1 products obtained by PEX reactions with nucleotides 6 a–6 c and further proved the stability of the acetonide protecting groups under elongation reaction conditions (Figure S22). By application of this protocol we also investigated the nature of the products stemming from reactions carried out with CTP 13 a. This analysis confirms that n+2 product is formed and the blocking group partially hydrolysed (Figure S23). After confirmation by MALDI-TOF of the formation of the targeted ssRNA n+1 products, we tested different conditions aiming at removing the acetal protecting groups. Deprotection of O,O-isopropylidene and -benzylidene protecting groups is usually accomplished under acidic conditions.42 Unlike DNA, RNA is much more resistant to acidic hydrolysis and the ribophosphodiester linkages are believed to be most stable at pH 4–5.43 Hence, we screened several acidic deprotection conditions compatible with acetal removal by incubating the unreacted RNA primer and analysing the integrity of the oligonucleotide by gel electrophoresis (Figure S24). Even though most conditions led to degradation of the oligonucleotide, some conditions appeared to be compatible with RNA. When we applied these conditions to the n+1 product obtained by PEX reaction with nucleotide 6 a, only incubation with 2 % DCA in H2O led to partial deprotection of the isopropylidene moiety (Figure 6). We also investigated other, less common, deprotection methods such as incubation with BCl3 12 % in MeOH or ZrCl4 (10 mol%) but these did not permit removal of the 2’,3’-isopropylidene protecting group and mainly led to degradation of the oligonucleotide (Figure S25A). A similar outcome was observed when n+1 products obtained with nucleotides 6 c and 6 d were treated with an acidic Dowex resin or neat pyrrolidine for 1 h at 50 °C (Figure S25B).
MALDI-TOF analysis of the deprotection of RNA primer elongated with a single 6 a nucleotide. The deprotection conditions involved incubation for 2.5 h in the presence of 2 % DCA in H2O at rt.
3 Discussion
Production of nucleic acids by de novo enzymatic synthesis is challenging for DNA and even more arduous for sugar modified substrates such as RNA or XNAs. The major difficulties reside in the identification of suitable and matching nucleotide/polymerase couples. Indeed, the protecting groups present on the nucleotide needs to be sufficiently stable to prevent hydrolysis from occurring during both enzymatic synthesis and upon storage but concomitantly must be labile enough to facilitate mild and rapid deprotection after incorporation. Moreover, polymerases need to obey the strict requirement of tolerating sugar-modified residues and incorporate single nucleotides with high efficiency and low coupling times. Preferentially, polymerases should display template-independent activity. Here, we have explored the possibility of using cis-diol protecting groups in conjunction with the template-independent RNA polymerases PUP and PAP. While 2’,3’-O-acetals are not yet the ideal blocking groups, they present a number of fundamentally important characteristics for efficient enzymatic de novo RNA synthesis. Indeed, pyrimidine nucleotides equipped with 2’,3’-O-isopropylidene moieties are excellent substrates for the commercially available PUP polymerase and allow for the high yielding production of extended primers within minutes and without formation of any side-products. The protected nucleotides are also stable upon storage in aqueous or buffered solutions. In addition, deprotection can be achieved by application of acidic conditions under which RNA is noticeably stable. These conditions are also compatible with certain base-modifications (e. g. N-methyl-pseudouridine) or phosphate alterations (e. g. phosphorothioates) which are key constituents of mRNA vaccines and therapeutic siRNAs, respectively.1b In addition, the simultaneous blocking of both hydroxyl moieties prevents the formation of undesired 2’-5’-branched RNA.
We found that the hydrolytic stability of the 2’,3’-O-isopropylidene moiety installed on nucleotides and during enzymatic reactions follows the order U > C > A, G, I. Hence, modulation of the substitution pattern of the isopropylidene scaffold (e. g. by introducing electron-withdrawing moieties) or alternate deprotection conditions are expected to favour completion of the deprotection reactions for pyrimidines, uridine particularly. On the other hand, purine nucleotides equipped with 2’,3’-O-isopropylidene masking groups are not well tolerated by any of the template-independent polymerases and appear to be partially hydrolysed during enzymatic synthesis. This low substrate tolerance might originate from a deviation from the C3′-endo sugar pucker44 and might be remediated by using engineered polymerases.17 Alternatively, a different protecting group strategy can be used for purine (e. g. 3’-O-allyl17) and pyrimidine (e. g. 2’,3’-O-isopropylidene) nucleotides to ensure efficient enzymatic incorporation of all ribonucleotides.
4 Conclusions
Herein, we have explored the possibility of using cis-diol protected nucleotides for PUP/PAP-mediated de novo RNA synthesis. Amongst the protecting groups that were screened, 2’,3’-O-isopropylidene offers a good compromising between robustness, bulkiness, polymerase recognition, and deprotection conditions at least for pyrimidine nucleotides. We have indeed demonstrated that equipping UTP and CTP with such a temporary masking group, permits the high yielding and efficient production of the RNA primers extended by a single nucleotide. Even though pyrimidines presenting such a modification pattern are not ideal candidates, these results bode well for the identification of potent blocking groups for enzymatic de novo RNA synthesis. Alternative cis-diols, including cyclic phosphate and silyl ethers, combined with engineered polymerases are currently investigated in our laboratory to favour single incorporation events of blocked nucleotides, particularly of purines, and identify mild and efficient deprotection conditions.
Experimental Section
General Protocols for Triphosphorylation of Blocked Nucleosides
Method A (Ludwig-Eckstein)
The suitably protected nucleoside (1 eq.) was dissolved in a pyrdine:dioxane 2 : 1 mixture (1.5 mL). The reaction was cooled down to 0 °C. Salicyl chlorophosphite (1.5 eq.) was then added portion wise under Argon atmosphere. The reaction was stirred for 4 h at 0 °C. After this time, tris(tetra-n-butylammonium) hydrogen pyrophosphate (1.5 eq.) was diluted in DMF (1 mL), this freshly prepared solution was then added to the mixture. Tributylamine (1.5 eq.) was added dropwise. The reaction was stirred for a further 1.5 h. Iodine (1.5 eq.) was dissolved in a pyridine/H2O 95 : 5 mixture (1 mL), this solution was added dropwise to the mixture. After an additional 30 min, the reaction was quenched by addition of Na2S2O3 sat. solution until disappearance of yellow coloration. The reaction was evaporated to dryness. The crude was dissolved in a minimum amount of H2O. The crude solution was poured dropwise to a 2 % NaClO4 in acetone solution for precipitation. Eppendorfs were centrifuged for 10–15 min. The acetone was poured out slowly and the white precipitate was dried under high-vacuum.
A) The crude was then purified by HPLC on anion exchange column DNAPacTM PA-100 BioLCTM (Thermo Scientific), 22 x 250 mm, Flow: 10 mL/min; gradient 0 % B for 5 min then 0 to 100 % B in 25 min and 100 % B for 5 min (A: 10 mM TEAB; B: 1 M TEAB) at rt.
B) The collected fraction was then lyophilized and purified again by HPLC using C18 column Kinetex (Phenomenex) 250 x 10.0 mm, 5 μM, 100 Å, Flow: 2 mL/min; Gradient: 0 to 50 % of B in 30 min (A: 20 mM TEAA, B: MeCN) at rt. Fractions of interest were lyophilized giving the desired pure 5’-triphosphate as a white foam.
Method B (Yoshikawa Protocol)
The suitably nucleoside (1 eq.) was solubilised in trimethylphosphite (c=0.25 M) at 0 °C. Freshly distilled phosphorous oxychloride was added dropwise (1.1 eq.). The reaction was stirred for 4 h at 0 °C under argon atmosphere. Bis-tributylammonium pyrophosphate (1.1 eq.) was dissolved in dry DMF. This previous solution and TBA (1.1 eq.) were added dropwise to the reaction mixture at 0 °C. The reaction was stirred for 1 h at 0 °C. The reaction was quenched by addition of TEAB 1 M solution and let stirred for 30 min at room temperature. The reaction was evaporated to dryness, co-evaporated with water. The crude was dissolved in a minimum amount of H2O. The crude solution was poured dropwise to a 2 % NaClO4 in acetone solution for precipitation. Eppendorfs were centrifuged for 10–15 min. The acetone was poured out slowly and the white precipitate was dried under high-vacuum. This solid was then purified by HPLC on anion exchange column DNAPacTM PA-100 BioLCTM (Thermo Scientific), 22×250 mm, Flow: 10 mL/min; gradient 0 % B for 5 min then 0 to 100 % B in 25 min and 100 % B for 5 min (A: 10 mM TEAB; B: 1 M TEAB) at rt.
General protocol of PUP/PAP-mediated extension reactions: RNA primer (20 pmol) is incubated with the modified nucleoside triphosphates (at a given concentration) with a metal cofactor and the PUP or PAP polymerase (10 U) in 1X reaction buffer (supplied with the polymerase; 10 μL final volume) at 37 °C for indicated reaction times. The reaction mixtures were then purified by Nucleospin columns and quenched by the addition of an equal volume of loading buffer (formamide (70 %), ethylenediaminetetraacetic acid (EDTA, 50 mM), bromophenol (0.1 %), xylene cyanol (0.1 %)). The reaction products were then resolved by electrophoresis (PAGE 20 %) and visualized by phosphorimager analysis.
Synthetic Procedures and Nucleotide Characterization2′,3′-Bis-O-acetyluridine-5′-triphosphate (4 a)
Triphosphorylation method B was followed, starting from 85 mg (0.259 mmol) of 3 a. After HPLC purification, the nucleotide 4 a was obtained in a very low yield (1.5 mg, 0.003 mmol, 1 %). The quantity was estimated by UV measurement taking the ϵ value of UTP as a reference (9.8 L mmol−1 cm−1). Data are in accordance with those reported in the literature.21b
1H NMR (500 MHz, D2O): δ 7.91 (d, J=8.12 Hz, 1H), 6.16 (d, J=4.94 Hz, 1H), 5.98 (d, J=8.10 Hz, 1H), 5.51-5.46 (m, 2H), 4.55-4.51 (m, 1H), 4.33-4.28 (m, 1H), 4.26-4.20 (m, 1H), 2.17 (s, 3H), 2.11 (s, 3H).
31P NMR (202 MHz, D2O): δ -10.93 (d, J=19.3 Hz, 1P), −11.66 (d, J=19.9 Hz, 1P), −23.26 (t, J=20.2 Hz, 1P)
HRMS (ESI) m/z [M−H]− calcd for C13H19N2O17P3 566.9824; Found 566.9828.
2′,3′-Bis-O-benzoyluridine-5′-triphosphate (4 b)
Triphosphorylation method A was followed starting from 62 mg (0.137 mmol) of 3 b. Product 4 b was purified by 1) anion exchange followed by 2) C18, affording 21 mg (0.030 mmol, 22 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 8.09 (d, J=8.12 Hz, 1H), 7.98 (d, J=8.44 Hz, 2H), 7.79 (d, J=8.44 Hz, 2H), 7.67 (t, J=7.46 Hz, 1H), 7.59 (t, J=7.48 Hz, 1H), 7.47 (t, J=7.84 Hz, 2H), 7.35 (t, J=7.86 Hz, 2H), 6.44 (d, J=5.96 Hz, 1H), 6.07 (d, J=8.12 Hz, 1H), 5.92 (dd, J=5.66, 3.66 Hz, 1H), 5.84 (t, J=5.86 Hz, 1H), 4.81-4.84 (m, 1H), 4.37-4.52 (m, 2H).
31P NMR (202 MHz, D2O): δ -10.91 (d, J=19.7 Hz, 1P), −11.66 (d, J=19.7 Hz, 1P), −23.27 (t, J=17.8 Hz, 1P).
13C NMR (125 MHz, D2O): δ 166.9, 166.6, 166.0, 151.4, 141.9, 134.4, 134.3, 129.6, 129.5, 128.8, 128.7, 128.2, 127.7, 103.1, 87.0, 81.6, 74.1, 71.9, 65.2.
HRMS (ESI): m/z [M−H]− calcd for C23H23N2O17P3 691.0137; Found 691.0137.
2′,3′-Bis-O-pivaloyluridine-5′-triphosphate (4 c)
Triphosphorylation method A was followed starting from 50 mg (0.121 mmol) of 3 c. Product 4 b was purified by 1) anion exchange only affording 3.2 mg (0.005 mmol, 4 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 7.89 (d, J=8.15 Hz, 1H), 6.12 (d, J=6.63 Hz, 1H), 5.95 (d, J=8.12 Hz, 1H), 5.45 (dd, J=5.20, 2.74 Hz, 1H), 5.38 (t, J=6.02 Hz, 1H), 4.53-4.49 (m, 1H), 4.27-4.22 (m, 1H), 4.21-4.15 (m, 1H), 1.20 (s, 9H), 1.09 (s, 9H)
31P NMR (202 MHz, D2O): δ -9.88 (d, J=20.1 Hz, 1P), −11.5 (d, J=18.4 Hz, 1P), −22.2 (t, J=18.4 Hz, 1P).
13C NMR (125 MHz, D2O): δ 180.2, 179.9, 179.8, 166.0, 151.7, 141.3, 103.2, 86.1, 81.9, 81.8, 73.6, 71.6, 38.6, 26.3, 26.2, 22.5.
HRMS (ESI): m/z [M−H]− calcd for C19H31N2O17P3 651.0763; Found 651.0768.
2′,3′-O-Isopropylideneuridine-5’-triphosphate (6 a)
Triphosphorylation method A was followed starting from 150 mg of 5 a (0.528 mmol). Product 6 a was purified by 1) anion exchange and then 2) C18 affording 44.0 mg (0.084 mmol, 16 %) of the desired nucleotide.
Triphosphorylation method B was followed starting from 74 mg of 5 a (0.297 mmol). Product 6 a was purified by 1) anion exchange only affording 10.0 mg (0.019 mmol, 6 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 7.69 (d, J=7.96 Hz, 1H), 5.85 (d, J=3.31 Hz, 1H), 5.77 (d, J=7.94 Hz, 1H), 4.95 (dd, J=6.25, 2.62 Hz, 1H), 4.89 (dd, J=6.22, 3.33 Hz, 1H), 4.46-4.42 (m, 1H), 4.11-4.07 (m, 2H), 1.45 (s, 3H), 1.28 (s, 3H).
31P NMR (202 MHz, D2O): δ -11.01 (d, J=20.1 Hz, 1P), −11.98 (d, J=20.1 Hz, 1P), −23.43 (t, J=20.1 Hz, 1P)
13C NMR (125 MHz, D2O): δ 179.4, 166.3, 151.4, 142.3, 114.4, 101.9, 92.6, 85.0, 84.9, 84.3, 80.8, 65.8, 65.7, 26.1, 24.3.
HRMS (ESI): m/z [M−H]− calcd for C12H19N2O15P3 522.9926; Found 522.9930.
2′,3′-O-Cyclohexylideneuridine-5’-triphosphate (6 b)
Triphosphorylation method A was followed starting from 95 mg of 5 b (0.151 mmol). Product 6 b was purified by 1) anion exchange only affording 8.2 mg (0.015 mmol, 5 %) of the desired nucleotide.
Triphosphorylation method B was followed starting from 75 mg of 5 b (0.231 mmol). Product 6 b was purified by 1) anion exchange only affording 6.6 mg (0.012 mmol, 5 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 7.87 (d, J=8.11 Hz, 1H), 5.98-5.94 (m, 2H), 5.10-5.05 (m, 2H), 4.65-4.61 (m, 1H), 4.28-4.18 (m, 2H), 1.77-1.61 (m, 6H), 1.61-1.51 (m, 2H), 1.51-1.37 (m, 2H).
31P NMR (202 MHz, D2O): δ −7.87 (d, J=18.6 Hz, 1P), −11.48 (d, J=18.8 Hz, 1P), −22.0 (t, J=19.1 Hz, 1H).
HRMS (ESI): m/z [M−H]− calcd for C15H23N2O15P3 563.0239; Found 563.0243.
2′,3′-O-Benzylideneuridine-5’-triphosphate (6 c)
Triphosphorylation method A was followed starting from 93.5 mg of 5 c (0.281 mmol). Product 6 c was purified by 1) anion exchange only affording 10.5 mg (0.018 mmol, 6 %) of the desired nucleotide as a mix of two diastereoisomers.
1H NMR (500 MHz, D2O): δ 7.93 (d, J=8.13 Hz, 0.4H), 7.87 (d, J=8.11 Hz, 0.6H), 7.69-7.65 (m, 0.8H), 7.63-7.60 (m, 1.2H), 7.57-7.50 (m, 3H), 6.26 (s, 0.6H), 6.14 (s, 0.4H), 6.13-6.10 (m, 1H), 5.98-5.94 (m, 1H), 5.26-5.18 (m, 2H), 4.68-4.64 (m, 0.6H), 4.38-4.27 (m, 2H).
31P NMR (202 MHz, D2O): δ −10.91 (d, J=19.5 Hz, 1P), −11.64 (d, J=19.9 Hz, 0.6P), −11.97 (d, J=20.0 Hz, 0.4P), −23.29 (t, J=19.7 Hz, 1P)
13C NMR (125 MHz, D2O): δ 166.4, 142.8, 1346, 130.5, 128.9, 128.8, 127.1, 127.0, 107.0, 103.6, 102.2, 101.8, 92.8, 91.5, 85.0, 84.1, 83.6, 82.6, 80.0, 58.6.
HRMS (ESI): m/z [M−H]− calcd for C16H19N2O15P3 570.9926; Found 570.9929.
2′,3′-O-Moc-Ethylideneuridine-5’-triphosphate (6 d)
Triphosphorylation method A was followed starting from 84 mg (0.256 mmol) of 11. Product 6 d was purified by 1) anion exchange and 2) C18 affording 25.4 mg (0.045 mmol, 17 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 7.77 (d, J=8.10 Hz, 0.7H), 7.71 (d, J=8.10 Hz, 0.3H), 5.86-5.79 (m, 2H), 5.59 (t, J=4.85 Hz, 0.3H), 5.41 (t, J=4.54 Hz, 0.7H), 4.98-4.91 (m, 2H), 4.57-4.54 (m, 0.7H), 4.43-4.39 (m, 0.3H), 4.20-4.09 (m, 2H), 3.65 (s, 2.1H), 3.63 (s, 0.9H), 2.89 (d, J=4.57 Hz, 1.4H), 2.81 (d, J=4.91 Hz, 0.6H).
31P NMR (202 MHz, D2O): δ −10.97 (d, J=19.6 Hz, 1P), −11.77 (dn J=19.8 Hz, 0.3P), −12.04 (d, J=20.0 Hz, 0.7P), −23.41 (t, J=19.8 Hz, 1P).
13C NMR (125 MHz, D2O): δ 179.4, 171.6, 166.3, 151.4, 142.8, 142.4, 103.9, 102.1, 101.8, 101.2, 92.6, 91.5, 84.6, 83.8, 82.3, 80.0, 65.84, 65.79, 52.5, 38.6, 38.4, 30.2.
HRMS (ESI): m/z [M−H]− calcd for C13H19N2O17P3 566.9824; Found 566.9820.
2’,3’-O-Isopropylidene-5’-triphosphate cytidine (12 a)
Triphosphorylation method A was followed starting from 70 mg (0.247 mmol) of 2’,3’-isopropylidene cytidine 11 a. Product 12 a was purified by 1) anion exchange and 2) C18 affording 1.9 mg (0.0037 mmol, 1.5 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 8.52 (s, 1H), 8.30 (s, 1H), 6.30 (d, J=3.43 Hz, 1H), 5.43 (dd, J=6.07, 3.45 Hz, 1H), 5.29 (dd, J=6.09, 1.99 Hz, 1H), 4.75-4.70 (m, 1H), 4.32-4.20 (m, 2H), 1.71 (s, 3H), 1.50 (s, 3H).
31P NMR (500 MHz, D2O): δ −10.89 (d, J=19.6 Hz, 1P), −11.76 (d, J=19.6 Hz, 1P), −23.25 (t, J=19.8 Hz, 1P).
13C NMR (125 MHz, D2O): δ 142.4, 114.2, 95.7, 93.2, 85.0, 80.8, 65.8, 26.2, 24.3.
HRMS (ESI): m/z [M-H]− Calcd for C12H20N3O14P3 522.0085; Found 522.00822’,3’-O-Isopropylidene-5’-triphosphate guanosine (12 b)
Triphosphorylation method A was followed starting from 70 mg (0.216 mmol) of 2’,3’-isopropylidene guanosine 11 b. Product 12 b was purified by 1) anion exchange and 2) C18 affording 23 mg (0.041 mmol, 19 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 8.05 (s, 1H), 6.04 (d, J=3.15 Hz, 1H), 5.33 (dd, J=6.06, 3.20 Hz, 1H), 5.21 (dd, J=6.08, 2.17 Hz, 1H), 4.56-4.60 (m, 1H), 4.15-4.22 (m, 2H), 1.62 (s, 3H), 1.42 (s, 3H).
31P NMR (500 MHz, D2O): δ −10.97 (d, J=19.4 Hz, 1P), −11.74 (d, J=19.3 Hz, 1P), −23.33 (t, J=19.2 Hz, 1P).
13C NMR (125 MHz, D2O): δ 178.8, 158.6, 153.8, 151.2, 137.7, 115.8, 114.8, 90.0, 84.7, 84.6, 83.7, 81.3, 65.7, 26.1, 24.4.
HRMS (ESI): m/z [M−H]− calcd for C13H20N5O14P3 562.0147; Found 562.0141.
2’,3’-O-Isopropylidene-5’-triphosphate adenosine (12 c)
Triphosphorylation method A was followed starting from 80 mg (0.260 mmol) of 2’,3’-isopropylidene adenosine 11 c. Product 12 c was purified by 1) anion exchange and 2) C18 affording 1.9 mg (0.0035 mmol, 1,3 %) of the desired nucleotide. Data are in accordance with those reported in the literature.45
1H NMR (500 MHz, D2O): δ 8.82 (s, 1H), 8.30 (s, 1H), 6.29 (d, J=3.44 Hz, 1H), 5.43 (dd, J=6.08, 3.44 Hz, 1H), 5.29 (dd, J=6.08, 2.00 Hz, 1H), 4–75-4.70 (m, 1H), 4.31-4.19 (m, 2H), 1.71 (s, 3H), 1.50 (s, 3H).
31P NMR (500 MHz, D2O): δ −10.89 (d, J=19.6 Hz, 1P), −12.06 (d, J=19.6 Hz, 1P), −23.30 (t, J=19.5 Hz, 1P).
13C NMR (125 MHz, D2O): δ 179.0, 154.2, 151.0, 148.6, 140.6, 118.4, 114.9, 90.4, 84.8, 84.7, 84.0, 81.4, 65.9, 26.2, 24.4.
HRMS (ESI): m/z [M−H]− calcd for C13H20N5O13P3 546.0198; Found 546.0200.
2’,3’-O-Isopropylidene-5’-triphosphate inosine (12 d)
Triphosphorylation method A was followed starting from 100 mg (0.32 mmol) of 2’,3’-isopropylidene inosine 11 d. Product 12 d was purified by 1) anion exchange and 2) C18 affording 6 mg (0.011 mmol, 3 %) of the desired nucleotide.
1H NMR (500 MHz, D2O): δ 8.41 (s, 1H), 8.22 (s, 1H), 6.30 (d, J=3.15 Hz, 1H), 5.44 (dd, J=6.02, 3.23 Hz, 1H), 5.28 (dd, J=6.10, 1.86 Hz, 1H), 4.72-4.68 (m, 1H), 4.27-4.17 (m, 2H), 1.67 (s, 3H), 1.46 (s, 3H).
31P NMR (500 MHz, D2O): δ −10.93 (d, J=19.5 Hz, 1P), −11.81 (d, J=19.8 Hz, 1P), −23.30 (t, J=19.6 Hz, 1P).
13C NMR (125 MHz, D2O): δ 179.1, 158.6, 148.5, 146.0, 140.0, 114.8, 90.8, 85.0, 84.1, 81.44, 65.8, 26.1, 24.3.
HRMS (ESI): m/z [M−H]− calcd for C13H19N4O14P3 547.0038; Found 547.0041.
Acknowledgments
The authors acknowledge generous funding from DARRI and Institut Carnot ‘Pasteur Microbes and Health’ Call 2021 (grant # INNOV-99-2, including a postdoctoral fellowship to M.P.). The pharmaceutical company Sanofi is acknowledged for funding a postdoctoral fellowship to C.K.. Patrick England and the Molecular Biophysics core facility of Institut Pasteur are acknowledged for their help with MALDI TOF acquisition. Dr. Luc Even and Dr. Jean Haensler are acknowledged for fruitful discussions.
Conflict of interests
The authors declare no conflict of interest.
