key: cord-0777969-4myxxsgf
authors: Shanmugasundaram, Muthian; Senthilvelan, Annamalai; Kore, Anilkumar R.
title: Recent Advances in Modified Cap Analogs: Synthesis, Biochemical Properties, and mRNA Based Vaccines
date: 2022-04-14
journal: Chem Rec
DOI: 10.1002/tcr.202200005
sha: a6afc5763d1ec31380b4c933c9d2c041baf96bdb
doc_id: 777969
cord_uid: 4myxxsgf

The recent FDA approval of the mRNA vaccine for severe acute respiratory syndrome coronavirus (SARS‐CoV‐2) emphasizes the importance of mRNA as a powerful tool for therapeutic applications. The chemically modified mRNA cap analogs contain a unique cap structure, m(7)G[5′]ppp[5′]N (where N=G, A, C or U), present at the 5′‐end of many eukaryotic cellular and viral RNAs and several non‐coding RNAs. The chemical modifications on cap analog influence orientation's nature, translational efficiency, nuclear stability, and binding affinity. The recent invention of a trinucleotide cap analog provides groundbreaking research in the area of mRNA analogs. Notably, trinucleotide cap analogs outweigh dinucleotide cap analogs in terms of capping efficiency and translational properties. This review focuses on the recent development in the synthesis of various dinucleotide cap analogs such as dinucleotide containing a triazole moiety, phosphorothiolate cap, biotinylated cap, cap analog containing N1 modification, cap analog containing N2 modification, dinucleotide containing fluorescence probe and TAT, bacterial caps, and trinucleotide cap analogs. In addition, the biological applications of these novel cap analogs are delineated.

The 5'-end of cellular and eukaryotic viral RNAs exhibits a matchless cap structure (m 7 G[5']ppp [5' ]N, where N is any nucleotide) that are synthesized by various RNA polymerases. [1] The N7 guanosine (m 7 G) residue in the messenger RNA (mRNA) is connected to the first nucleotide of the mRNA transcript via a triphosphate linkage between the two 5'hydroxyl groups as shown in Figure 1 . The distinct cap structure plays a vital role in providing resistance to 5'exonuclease that helps to protect the mRNA from speedy degradation by 5'-exonuclease activity. It has participated in numerous aspects of mRNA metabolism such as intracellular transport, splicing, subcellular localization, initiation of RNA, translation, and mRNA turnover. [2, 3] The most appealing feature of the cap is its strong ability to recognize eukaryotic initiation factor 4E (eIF4E) during the initiation of translation that has been overexpressed in several categories of tumors. [4] The presence of both the delocalized positive charge of the m 7 guanosine base moiety and the triphosphate bridge in the cap structure contributes to the specific recognition to eIF4E. [5] It has been acknowledged in the literature that eIF4E is a striking target for anticancer directed therapy. [4] The modified cap analog serves as a potential inhibitor to target eIF4E in which the cap analogs may neutralize the overexpressed levels of eIF4E that is highly depend on its ability to bind with eIF4E. In addition to eIF4E protein, the cap structure recognizes other cap-binding proteins such as CBC80/20 nuclear cap-binding complex and Vaccinia virus methyltransferase VP39. [6, 7] The most common strategy for the in vitro synthesis of 5'capped mRNAs comprises the use of synthetic dinucleotide analog, m 7 G[5']ppp [5' ]G (standard cap), as an initiation of transcription. However, the existence of two free 3'-OH groups on both guanosine moieties promotes as the initiating nucleophile for transcription elongation that results in both the forward (m 7 G[5']ppp [5' ]G[pN] n ) and the reverse orientation (G [5' ]ppp [5' ]m 7 G[pN] n ) products. [8] The mRNA transcripts with forward orientation product is recognized, whereas transcripts with reverse orientation product is not properly recognized during the translational process that decreases the overall translational properties. The breakthrough in the field of mRNA cap is the innovation of an anti-reverse cap analog (ARCA) that circumvents the problems encountered in the standard cap. [9, 10] The ARCA involves 3'-OH modification on m 7 G moiety in which the modified cap analog incorporates exclusively in the forward orientation due to the absence of the initiating nucleophile on m 7 G moiety. Moreover, the exclusive Muthian Shanmugasundaram received his Ph.D. in Organic Chemistry from the University of Madras, Chennai (India). He did three post-doctoral research at the Tsing Hua University, Hsinchu (Taiwan), University of Arkansas, Fayetteville (USA) and University of Texas at El Paso (USA). His doctoral and post-doctoral research work involved in the area of heterocyclic chemistry, metal-mediated organic synthesis, solid-phase organic synthesis, combinatorial chemistry, supramolecular chemistry, and medicinal chemistry. Currently, he is a Senior Staff Scientist at Thermo Fisher Scientific, Austin, Texas, USA. He has authored in more than 70 peer-reviewed research papers and has the inventorship of nine patents. His current research focuses in the area of nucleic acid chemistry, mRNA cap analogs for therapeutics, proteinase K inhibitors, and lipid-based transfection reagent.

Annamalai Senthilvelan received his Ph.D degree in Organic Chemistry (2002) from University of Madras-India, under Prof. V. T. Ramakrishnan in the area of Organic photochemistry and heterocyclic chemistry. He carried out two post-doctoral studies in Taiwan (2002) (2003) (2004) (2005) (2006) (2007) . One at Academia Sinica in the area of carbohydrate chemistry and another at National Chiao-Tung University in the area of Organic supramolecular host-guest chemistry. He joined Prof. Howard E. Zimmerman's group in the department of chemistry, University of Wisconsin-Madison as research associate (2007) (2008) (2009) (2010) 2'-OH modification on m 7 G moiety also serves as the antireverse cap analog even when there is a free 3'-OH group on m 7 G moiety. [11] In general, the use of ARCA provides more than two fold translational properties as compared to a standard cap. [12, 13] The sugar modifications on m 7 G moiety not only influences the nature of the orientation but also promotes the translational efficiency, binding affinity, and nuclease stability. Furthermore, the modification of triphosphate chain (i. e.) thiophosphoralate cap analogs display higher translational properties than their parent triphosphate analogs. [14] The formation of IVT mRNA transcripts from the use of dinucleotide cap analog (m 7 GpppN) results in cap 0 structure that does not contain the 2'-OMe group in the first transcribed nucleotide. The cap 0 structure can also be generated by the use of the Vaccinia virus complex that is accomplished through post-transcriptional modifications. [15] In the case of higher mammalian cells, the IVT mRNA transcripts with cap 0 structure can further undergo post-transcriptional modifications to cap 1 structure (m 7 GpppNm wherein, Nm is a 2'-Omethylated nucleotide) and occasionally to cap 2 structure (m 7 GpppNmNm) even though cap 0 structure has the ability to initiate mRNA translation. [16] The cap 0 structure has been converted into cap 1 structure that involves the transfer of a methyl group from S-adenosylmethionine by the use of an enzyme, 2'-O-methyltransferase. [17] However, this method is not practically feasible for large scale preparations. During viral replication, it is important for the immune system to differentiate between non-self and self RNA. The presence of the 2'-O-methyl group in the first transcribed nucleotide plays a vital role to identify self RNA in the innate immune system and discriminate non-self RNA that helps to suppress viral replication and pathogenesis. [18, 19] The limitations associated with cap 0 structure has been effectively addressed by the use of a trinucleotide cap analog. [20] Cotranscriptional capping with m 7 GpppNmN-derived trinucleotide generates the cap 1 structure, in which the first transcribed nucleotide has the 2'-O-methyl group. The synthetic trinucleotide cap analog is used to make the IVT mRNA transcripts of Pfizer-BioNTech mRNA vaccine for severe acute respiratory syndrome coronavirus (SARS-CoV-2). [21a] In the case of Moderna mRNA vaccine, IVT mRNA transcript containing cap 1 structure has been obtained by the enzymatic mRNA capping involving Vaccinia capping enzyme and Vaccinia 2'-O-methyltransferase. [21b] The importance and biological application of mRNA cap analog are well recognized through the publications of several review articles. [22, 23] Since our previous review in 2017, [24] an array of breathtaking research has emerged in the area of cap analogs, which warrants a new review article to deliver an update and guidance in this research topic. In this review, we summarize an overview of the literature for the chemical synthesis of several modified cap analogs and their biochemical properties, concentrating on advances in the last five years. To facilitate discussion, the chemical synthesis of cap analog has been classified into three different groups such as dinucleotide cap analogs, bacterial caps, and trinucleotide cap analogs. The biochemical properties such as capping efficiency, binding affinity, translational efficiency, and resistant to decapping enzyme for these three groups have been discussed in the biological application part. 

The multiple functional groups such as primary and secondary hydroxyl groups in sugar and amino group in base moiety of the nucleoside/nucleotide make the chemical synthesis of cap analogs extremely challenging. Moreover, these nucleosides/ nucleotides are known to cleave under acidic/basic conditions. The coupling reaction between activated nucleotide and nonactivated nucleotide is the most commonly used strategy for the synthesis of cap analogs. [23] The activated nucleotide acts as the electrophile, whereas the non-activated nucleotide acts as the nucleophile. While there are numerous activating groups such as phenylthio group, [25, 26] methoxyphenylthio group, [27, 28] 5-chloro-8-quinolyl group, [29] imidazolide, [30, 31] and morpholidate [9] have been utilized for the coupling reaction, the most practical and feasible activating group utilizes the use of imidazolide group. Although several types of a divalent metal catalyst such as MgCl 2 , MnCl 2 , CaCl 2 , CdCl 2 , and ZnCl 2 have been employed in both aqueous and anhydrous conditions, [32] the use of zinc chloride under anhydrous conditions involving polar aprotic solvent, DMF provides high yields of product with shortening of reaction time. [33] The involvement of lewis acid catalyst is crucial for the successful coupling reaction that brings both activated and non-activated nucleotide into close vicinity by divalent metal coordination (Scheme 1). [32] Although these two reactants as such are poorly soluble in DMF, the lewis acid catalyst acts as a template that makes the two reactants solubilize under DMF conditions. The pre-requisite for the non-activated nucleotide is the presence of hydrophobic counter ion and the most commonly used one is triethylammonium ion. In contrast to the silica gel purification for the organic compounds, the purification for a cap analog employs the use of anion exchange column chromatogrphy. The most frequently used anion exchange resin is diethylamino ethyl (DEAE) Sepharose or Sephadex and eluting buffer is 1.0 M triethylammonium bicarbonate buffer (TEAB). The total number of phosphate groups present in the mixture dictates the order of elution during the purification process. In general, the monophophate elutes first, then diphosphate and finally cap analog during elution. The chemical synthesis of triphosphate non-bridging cap analog results in a mixture of two diastereomers. These two diasteromers are separated and isolated as a single isomer by anion exchange followed by reverse phase chromatography using Supelcosil LC-18-T column.

Walczak and co-workers reported a new route for synthesizing 5' cap mimics and capped RNAs and studied its biochemical properties. [34] The copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC) reaction of nucleotide containing a terminal alkyne within the phosphate chain 11 with azido derivative of nucleoside 12 in several combinations in the presence of copper sulphate and sodium ascorbate under aqueous conditions afforded the corresponding dinucleotide cap analogs containing a triazole moiety 1-6 (Scheme 2). The type 1 class represents a dinucleotide cap containing a triazole moiety directly attached to guanosine moiety, whereas type 2 represents a dinucleotide cap containing a triazole moiety directly attached to m 7 G moiety. Similarly, the CuAAC reaction of nucleotide containing a terminal alkyne within the phosphate chain 13 with azido derivative of nucleotide 14 in different combinations under standard CuAAC reaction conditions furnished the corresponding dinucleotide cap analog containing a triazole moiety within the phosphate chain 7-10 (Types 3 and 4 -Scheme 3). The crude reaction mixture was quenched with EDTA in order to remove the copper ion and there were 34 dinucleotide cap analogs synthesized and isolated in high purity by using reverse phase HPLC purification (Schemes 2 and 3).

Mamot and co-workers reported a novel synthesis of dinucleotide cap analogs containing a functionalized azide group at the 2'-O or 3'-O position of m 7 G moiety in order to study the site-specific labelling of mRNA in living cells (Scheme 4). [35] The key intermediate, Im-m 7 GMP containing a functionalized azide group 21 was obtained in three steps, starting from GMP 18. The chemical synthesis of a series of dinucleotide cap analogs containing a 5'-phosphorothiolate (5'-PSL) moiety has been reported by Wojtczak et al. in order to study their biochemical properties such as translational efficiency, capbinding interactions and decapping assays (Schemes 5 and 6). [36] There are two different types of synthetic approaches that were utilized to make dinucleotide cap analogs containing a 5'-PSL moiety. First, the synthetic strategy involves the use of S N 2 S-alkylation reaction. The nucleophilic reaction of 7methylguanosine containing a 5'-thiophosphate (27 or 28) with 5'-iodo guanosine (29) in the presence of DBU 26 as a base, using DMSO as a solvent afforded the corresponding dinucleotide cap analogs containing a α-PSL modification 30 and 31, respectively in moderate yields (Scheme 5). Similarly, treatment of guanosine containing a 5'-thiophosphate (32, 33, or 34) with 5'-iodo 7-methylguanosine (35 or 36) afforded the corresponding dinucleotide cap analogs containing a γ-PSL modification 37-40, respectively. Second, the synthetic approach utilized the use of conventional coupling reaction between activated and non-activated nucleotide. Treatment of modified m 7 GDP (27, 41, or 42) with 5'-S-ImGMP 44 under standard ZnCl 2 /DMF system afforded the corresponding dinucleotide cap analogs containing a α-PSL modification 45-48, respectively in moderate yields. The coupling reaction of 5'-S-GDP (49 or 50) with modified Im-GMP (25, 43 a, or 43 b) in the presence of ZnCl 2 as a catalyst and DMF as a solvent furnished the corresponding cap analogs containing a γ-PSL modification 51-55, respectively (Scheme 6). The first approach has been used for the synthesis of dinucleotide cap analogs containing the exclusive α-PSL or γ-PSL modifications, whereas the second approach has been utilized for the synthesis of dinucleotide cap containing two or three modifications such as α-PSL, γ-PSL, and β-PS.

Bednarek et al. reported a series of biotinylated dinucleotide cap analogs with phosphate modification to study the effect of biochemical properties such as capping efficiency, translation efficiency, and hDcp2 decapping susceptibility. [37] The synthetic pathways leading to the formation of biotinylated dinucleotide cap anlogs 66-70 were depicted in Schemes 7 and 8. The key intermediate, 7-methylguanosine containing biotinylated moiety 60 for the synthesis of 66 and 68-70 was obtained in three steps, starting from 2'-amino-2'deoxyguanosine (56) . The monophosphorylation of 56 using phosphorous oxychloride as a phosphorylating agent, followed by the methylation of the resulting 57 using methyl iodide as a methylating agent, and subsequent biotinylation of the resulting 58 using biotin N-hydroxysuccinimide ester (59) under aqueous conditions furnished the corresponding 7methylguanosine containing a biotinylated moiety 60. The final coupling reaction of 7-methylguanosine containing a biotinylated moiety 60 with P-imidazolides of GDP 65 a, or methylenebisphosphonate 65 b in the presence of ZnCl 2 /DMF system afforded the corresponding biotinylated dinucleotide cap analogs 66 and 70, respectively. Compound 60 was activated into imdazolide salt 61, which was further coupled with guanosine diphosphate containing phosphate modifications 63 or 64 under standard conditions to afford the corresponding cap analogs 69 and 70, respectively. The imidazole activated-GMP 61 was coupled with thiophosphate and subsequent reaction of the resulting biotinylated thio Walczak and co-workers reported the chemical synthesis of several phosphotriazole cap analogs in order to study its biochemical properties of binding affinity for eIF4E, susceptibility to hDcp2-catalyzed decapping and translational efficiency. [38] The copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC) reaction of phosphonate derivatives of 7,2'-O- The chemical synthesis of a new series of dinucleotide cap analogs containing an alkyne handle at the N1-position of guanosine has been reported by Kopcial and co-workers in order to study the effect of cap-binding proteins. [39] To begin the synthesis, the key intermediate P-imidazolide 83 was obtained in three steps starting from guanosine. The reaction between guanosine (80) and propargyl bromide in the presence of sodium hydride and tetrabutylammonium iodide afforded the corresponding N1-propargylguanosine (81). The monophosphorylation of 81, followed by the activation of the resulting monophosphate 82 by using imidazole/triphenylphosphine/aldrithiol system afforded the desired imidazolide 83. Compounds m 7 GDP, m 7 GpCH 2 P, or m 7 GpNHp were coupled with P-imidazole 83 in the presence of zinc chloride as a catalyst and DMF as a solvent afforded the corresponding cap analog with unmodified phosphate bridge 84 and cap Piecyk and co-workers reported a chemical synthesis of novel double functionalized dinucleotide cap analog containing fluorescence probe and peptide moiety in order to explore the possibility of utilizing the cap analog for labelling purpose. [40] The coupling reaction of double functionalized Kocmik et al. reported a chemical synthesis of new dinucleotide cap analogs containing N2 modifications on m 7 G moiety in order to study its biochemical properties. [41] The coupling reaction of N7-methyl guanosine monophosphate containing N2 modification 108 with Im-GDP 109 in the presence of zinc chloride as a catalyst furnished the corresponding dinucleotide cap analog containing N2 modification on m 7 G moiety 110 (Scheme 15).

Piecyk and co-workers reported a chemical synthesis of novel dinucleotide cap analog containing transactivator of transcription (TAT) and dinucleotide cap analog containing TAT and dansyl dye in order to study the effect of human immunodeficiency virus 1 (HIV-1) TAT derived peptide. [42] Treatment of guanosine monophosphate containing N2 modification 111 with Im-m 7 GDP 103 in the presence of zinc chloride as a catalyst afforded the corresponding dinucleotide cap analog containing N2 modification 112. The reaction between 112 and N 3 -TAT under standard click chemistry reaction conditions afforded the corresponding dinucleotide cap analog containing TAT 113 (Scheme 16). Treatment of dinucleotide cap analog containing dansyl dye 106 with N 3 -TAT under standard click chemistry reaction conditions furnished the corresponding dinucleotide cap analog containing TAT and dansyl dye 114 (Scheme 17).

Nicotinamide adenine dinucleotide (NAD) is a coenzyme that consists of two nucleotides joined by pyrophosphate group. [43, 44] NAD-capped RNAs has been termed as bacterial caps that display some structural analogy to the eukaryotic mRNAs containing 7-methylguanosine cap present at the 5'end. [45] Although NAD has been known for over 100 years, the molecular biological applications of NAD-capped RNAs has been under-explored area of research. In this context, Mlynarska-Cieslak and co-workers reported the chemical synthesis of a series of NAD cap analogs in order to study the susceptibility of NAD-RNAs with deNADding enzymes. [46] The required intermediate, imidazole-activated nicotinamide mononucleotide 117 a was obtained in two steps, starting from the commercially available NAD 115. The hydrolysis of 115 in the presence of aqueous ZrCl 4 afforded nicotinamide riboside (NR) 5'-monophosphate 116. Activation of 116 using imidazole/aldrithiol/triphenylphosphine system afforded the corresponding P-imidazole 117 a. The coupling reaction of adenosine 5'-phosphorothioate (117 b) with P-imidazole 117 a in the presence of ZnCl 2 /DMF system afforded the corresponding two stereoisomers of NAD containing α-phosphor- 

Sikorski and co-workers reported a series of trinucleotide cap analogs and studied protein expression of exogenously delivered mRNA in three different mammalian cell lines and their susceptibility to decapping. [47] The required starting material, dinucleotide 5'-phosphate 129 (pNpG) was synthesized using solid-supported standard phosphoramidite chemistry. The coupling reaction of various dinucleotide 5'-phosphate 121, 123, or 129 with Im-m 7 GDP 103 under traditional ZnCl 2 / DMF system afforded the corresponding trinucleotide cap analog 130 (Scheme 20).

The molecular biological application of locked nucleic acid (LNA) technology serves as a powerful tool in the research areas such as geneotyping and antisense and antigene 

therapies. [48] The unique conformationally restricted structural feauture provides nuclear resistance, increased thermal stability, hybridization specificity, and sequence stability. [49] We envisaged that the design of a new trinucleotide cap analog bearing an LNA moiety could act as a potential molecular biology tool in the area of mRNA transfection applications such as anticancer immunization, protein production, and gene therapy and mRNA-based vaccines. In this context, we have reported a synthesis of novel trinucleotide cap analog bearing an LNA moiety and studied its capping efficiency and translational efficiency. [50] The required intermediate, 

acetonitrile as an activator, followed by the oxidation using iodine/pyridine mixture afforded the corresponding protected 

The capping efficiency assay provides a powerful tool to determine whether the newly synthesized cap is a substrate for RNA polymerases such as T7, SP3, and SP6 or not. [22f] Furthermore, this assay helps to determine the nature of the orientation as evidenced by the HPLC analysis of the RNA transcription mixture containing short oligonucleotides. [13d] The chemical modifications of cap analogs influence the outcome of the capping efficiency. In addition, factors such as ionic strength and ratio of cap analog to nucleotide mixture determine the capping efficiency of cap analogs. It has been revealed in the literature that eIF4E protein serves as a powerful weapon for anticancer directed therapy. [4] The cap-dependent translation is the major contributing causes to tumorigenesis. The binding of the 5' terminal mRNA cap structure with eIF4E is the rate limiting step of cap-dependent translation and the concentration of eIF4E influences the regulatory nexus of translation. The proper design of new chemically modified synthetic cap analog can be utilized as an inhibitor that binds with the elevated eIF4E levels in tumor cells and block the function of eIF4E.

The stability against nuclease plays an important role for the positive therapeutic application based on nucleotide drug. The 5'-terminal cap structure provides a powerful strategy to enhance nuclease stability and protect cytoplasmic mRNAs. However, several decapping enzymes are involved to remove the cap structure from RNA that results to disease development. [51, 52] Consequently, it is vital to selectively regulate the mRNA gene expressions that remain to be challenging clinical problems. There are two major mRNA decay pathways that are involved in eukaryotic cells. [53] First, initiation of shortening of the 3' poly(A) tail by the PAN2/3 and CCR-NOT complexes. Second, mRNA undergoes degradation by 5'!3' hydrolysis due to exonuclease and 3'! 5' hydrolysis due to exosome. In eukaryotic cells, DcpS belongs to the HIT family and Dcp1-Dcp2 complex belongs to the Nudix family which are the two types of decapping enzymes. The biochemical degradation studies of capped mRNA by scavenger decapping enzyme, DcpS, results in the formation of m 7 GMP suggesting that the site of cleavage is between the βand γ-phosphate bond of the cap moiety. In the case of Dcp1-Dcp2 complex, capped mRNA results in the formation of m 7 GDP and 5'-monophosphate RNA that suggests cleavage is happened between the αand βphosphate bond of the cap moiety. (Figure 2 ). The chemically modified mRNA cap analogs play a major role to provide resistance to hydrolysis by decapping enzymes that supports to prevent mRNA against degradation and enhance protein production.

The biochemical properties such as capping efficiency, translation efficiency, and susceptibility to decapping of 5'capped with analogs 15-17 are summarized in Table 1 . [35] The capping efficiency for new analogs 15-17 ranges from 72 to 85 %, indicating that these analogs are good substrates for SP6 RNA polymerase (entries 4-9). These analogs display slightly lower efficiency than reference cap analogs (entries 1 and 3). The mRNAs capped with analogs 17 a-d are translated up to 7.5-fold higher than those for mRNAs capped with standard cap and up to 4-fold higher than those capped with ARCA (entries 1, 2, and 6-9). The highest translational efficiency is observed for cap analog, 2'-N 3 -m 7 Gpp s pG D1 17 a (entry 6). The replacement of oxygen to sulphur at β-position in the triphosphate bridge provides up to 2.3-fold higher translational efficiency as compared to the parent compound (entries [4] [5] [6] [7] [8] [9] . These data are in agreement with the previously reported increased translation efficiency provided by the β-phosphorothioate group. [14] The βphosphorothioate modification influences the susceptibility of RNA to decapping that provides stabilization of the transcripts which is comparable to that of m 2 7,2'À O Gpp s pG D2-capped RNA (Table 1 , entries 3 and 6-9).

The susceptibility of the 5'-phosphorothiolate cap analogs to hydrolysis by the human DcpS enzyme was determined and the results are compiled in Table 2 . [36] The cap analogs containing α-PSL modification undergoes hydrolysis by DcpS  (entries 2, 3, 8, 11, and 12) . It is noteworthy that the cap analogs containing γ-PSL modification are resistant to hydrolysis by DcpS (entries 4-7, 9, 10, and 13-16). It appears that the replacement of 5'-O next to the m 7 G moiety with 5'-S protects the cap structure efficiently and prevents from mRNA degradation. These results are in similar agreement with the previously reported for O-to CH 2 substitution at the β-γ bridging position and several other modifications within the βor γ-phosphate. [54, 55] It is interesting to compare the results between susceptibility of hDcpS data with translational property data. The mRNAs capped with analogs 40 and 45 in HeLa cells provide 1.1-fold and 1.4-fold higher translational efficiency, respectively as compared to ARCA. [36] Although mRNAs with analog 40 at their 5'-end is more stable against hDcpS than mRNAs with analog 45, the translational efficiency of analog 40 ( fold more than clinically tested RG3039 (EC50 = 6.82 � 1.08 nM). [36] The biochemical properties such as capping efficiency, translation efficiency, and susceptibility to decapping of 5'capped with analogs 66-70 are compiled in Table 3 . [37] The capping efficiency for new biotinylated cap analogs 66-70 are ranged from 39 to 64 %, indicating that these analogs are substrates for SP6 RNA polymerase (entries 4-8). The mRNAs capped with biotinylated cap analogs 66-70 result in comparable or lower translational efficiency than standard cap analog (entries 1 and 4-8). These data indicate that the presence of 2'-substituted biotin moiety in cap analog negatively influence the overall translational efficiency of capped mRNA. In order to study the susceptibility of biotinlabelled cap analogs, short RNAs were subjected to enzymatic degradation by hDcp2. It is noteworthy that phosphate modified cap analogs such as imidodiphosphate 69 and methylenebis(phosphonate) 70 introduced at α,βposition of triphosphate provide resistance to mRNA degradation, whereas imidodiphosphate introduced at β,γ-position of triphosphate 67 undergoes slow mRNA degradation (entries 5, 7 and 8). The β-S-modified cap 68 provides partial resistant to mRNA degradation (entry 6). The binding affinity constants (K AS ) of biotin-labelled cap analogs 66-70 for eIF4E are comparable to that of the standard cap analog, indicating that biotin moiety's presence does not impact cap-eIF4E interaction to a considerable degree and the specificity of binding is maintained.

The biochemical properties such as capping efficiency, translation efficiency, and binding affinity to eIF4E with analogs 76 and 79 are summarized in Table 4 . [38] The capping efficiency for new analogs 76 a-d are ranged from 30 to 42 %, indicating that cap analogs containing a triazole moiety located between phosphate chains are not efficiently recognized by RNA polymerase (entries 4-7) . The capping efficiency of compound 79 provides 77 % efficiency (entry 8). These results show that cap analog containing a triazole moiety directly attached to the sugar is more efficiently recognized by RNA polymerase than cap analog containing a triazole moiety located between phosphate chains (entries [4] [5] [6] [7] [8] . The mRNAs capped with analogs 76 a-d provide higher translational efficiency up to 2.8-fold as compared to standard cap analog (entries 2 and 4-7), but mRNA capped with analog 79 provide 6.7-fold lower translational efficiency as compared to standard cap analog (entries 2 and 8). It seems that a triazole moiety directly attached to m 7 G in cap analog provides unfavorable effect on the overall translational efficiencies (entry 8). In contrast, a triazole moiety located between two phosphate chains does not affect the translational data (entries 4-7). The K AS value for cap analogs containing a triazole moiety located between the phosphate chains 76 a-d provides up to 6.7-fold higher values as compared to standard cap analog, whereas cap analog containing a triazole moiety attached directly to m 7 G gives slightly lower than the standard cap analog (entries 2 and 4-8).

In order to study the binding effect of substitution at N1 position of cap with eIF4E, the binding constant values (K D ) for the cap analog 84-88 -eIF4E complexes were determined. [39] The data show that cap analogs 84-88 have binding affinities comparable to that of the standard cap analog. It seems that the propargyl moiety attached to the N1 position of guanosine does not have any major impact on the interaction with eIF4E. Under standard conditions, cap analogs modified within the phosphate bridge 85-88 are resistant to hDcpS. It is interesting to note that fluorescent probes 94 and 95 display high affinity ligands for DcpS in which probe 95 (K D = 14 nM) has 2.5-fold greater affinity for DcpS than probe 94 (K D = 38 nM). [39] The capping efficiency of various cap analogs containing a triazole moiety 1-10 show lower capping efficiency as compared to standard cap analog. [34] The mRNAs capped with cap analogs 1-9 result in lower translational efficiency (0.06 to 0.5) as compared to standard cap (1.0), whereas mRNA capped with cap analog 9 provides comparable translational efficiency (0.89 � 0.11) to that of standard cap (1.0). [34] Table 5 summarizes the translational properties of differently capped mRNA in rabbit reticulocyte lysate (RRL) and human embryonic kidney derived (HEK293) cell lines and also decapping by Dcp1/Dcp2 complex. [41] The mRNAs capped with analogs 110 a,110 b in RRL cell line provide lower translational efficiency as compared to ARCA cap analogs (entries 2-5). The highest translational efficiency of 1.72 is observed for cap analog, bn 2 m 2 7,2'À O GpppG 110 c in RRL cell line (entry 6). The mRNAs capped with analogs 110 a-c in HEK293 cell line provide up to 2.2-fold higher translational efficiency as compared to ARCA cap analog (entries 3-6). The decapping results show that the presence of N2 modification within guanine of 110 a-c decreases stability as compared to ARCA analog (entries 3-6).

The ability of cap analogs containing TAT to inhibit capdependent translation was determined by the in vitro translation of capped-Renilla luciferase mRNA. Both compounds 113 and 114 inhibit in vitro cap-dependent translation that strongly indicates the presence of peptide moiety attached to the cap analog does not impact the inhibitory properties. It is noteworthy that the cap analog containing fluorescence probe and TAT 114 translocate into the human breast adenocarcinoma cancer cell line MCF-7. [42] In a series of nicotinamide-containing cap analogs, all cap analogs 117-119 and 125-128 undergo incorporation into RNA to generate the corresponding modified NAD-capped RNAs. [46] Remarkably, mRNAs capped with nicotinamidecontaining trinucleotide cap analogs 128 a and 128 b are resistant to deNADing enzymes such as NudC, Nudt12, and DXO. [46] Table 6 summarizes the biochemical properties such as capping efficiency, translational efficiency, binding affinity to eIF4E, and susceptibility to hDcp2 for various trinucleotide cap analogs 130 a-j. [47] The capping efficiency of cap analogs containing a purine nucleotide at the position of the first [11] [12] [13] [14] .

The capping efficiency of cap analogs containing a pyrimidine nucleotide at the position of the first transcribed nucleotide is comparable to that of the standard cap and ARCA (entries 7-10, 13, and 14). The capping efficiency of cap analogs containing a purine nucleotide at the position of the first transcribed nucleotide provides higher capping efficiency than cap analogs containing a pyrimidine nucleotide at the position of the first transcribed nucleotide (entries 2-8 and [11] [12] [13] [14] . The nature of the base at the first transcribed nucleotide plays an important role on the outcome of the translational efficiency. The presence of adenine at the first transcribed nucleotide provides the highest translational efficiency as compared to other bases such as cytosine, uracil, and guanine (entries 2-5). The presence of guanine at the first transcribed nucleotide provides the lowest translational efficiency (entries 11 and 12) . The translational efficiency is in the order of A > C > U > G. It is interesting to compare the effect of 2'-OMe group of the mRNA transcripts with different cell lines such as JAWS II (mouse immortalized immature dendritic cells), 3T3-L1 (mouse embryonic cells) and HeLa (cervical cancer). The presence of 2'-OMe group at the first transcribed mRNA in a dendritic cell provides up to seven-fold higher transla-tional efficiency as compared to the unmethylated one (entries 2-5, 7, and 8). The presence of the 2'-OMe group at the first transcribed mRNA in 3T3-L1 provides comparable translational efficiency to that of the unmethylated ones (entries 2, 3, 7, and 8). The presence of the 2'-OMe group at the first transcribed mRNA in HeLa cell lines provides slightly higher translational efficiency to that of the unmethylated ones (entries 2, 3, 7, and 8). The highest translational efficiency is observed for m 7 Gppp m6 A m pG 130 d that is 7.39fold higher translational efficiency as compared to unmethylated one, m 7 GpppApG 130 a (entries 2 and 5). Given the binding affinities and susceptibility to decapping by hDcp2 data, it appears that the presence of the 2'-OMe group at the first transcribed mRNA does not influence cap-eIF4E interaction and susceptibility to decapping as compared to the unmethylated cap analog (entries [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] .

The capping efficiency of new LNA tricap 143 was performed in an in vitro transcription system by using an IVT template encoding GFP to produce transcripts of~1000 nucleotides in length with a synthetic 5'-UTR, a mouse alpha globin 3'-UTR, and a 120 nucleotide poly(A) tail. [50] The outcome of the assay reveals that LNA cap 143 has a capping efficiency of 53 %, indicating this LNA tricap is a good substrate for T7 RNA polymerase. Next, we determined the translational efficiency of the different caps in JAWS II cells that has been used as a model for antitumor or pathogenspecific immunity studies. The translational efficiency data of LNA tricap 143, standard trinucleotide cap analogue (GAG), m 7 G(5')ppp(5')A m pG and anti-reverse cap analogue (ARCA), m 2 7,3'À O G(5')ppp(5')G are compiled in Figure 3 . Notably, the 

mRNA capped with LNA tricap is translated 5 times higher than GAG trinucleotide or ARCA capped mRNA transcript in the dendritic line.

The effect of chemical modification on a cap analog has had a significant impact on the outcome of the biochemical properties such as capping efficiency, translation efficiency, binding affinity to eIF4E, and susceptibility to decapping. [24] The invention of trinucleotide cap analog fostered much research and considerable progress in the area of mRNA cap analogs.

The utilization of cotranscriptional m 7 GpppNmN-derived trinucleotide in the cotranscriptional reaction produces the corresponding IVT mRNA with cap 1 structure, whereas cotranscriptional capping with m 7 GpppN-derived dinucleotide results in the formation of the corresponding IVT mRNA cap 0 structure. The presence of cap 1 structure in mRNA plays an important role to promote translation and also allow cells to differentiate between self and non-self RNA and suppress viral replication. [56] Remarkably, the trinucleotide cap analog outperforms the dinucleotide cap analog in terms of capping efficiency and translational properties. Based on the remarkable translational properties of m 7 Gppp m6 A m pG 130 c and m 7(LNA) -G(5')ppp(5')( 2'-OMe A)pG 143, it is highly likely that these new trinucleotide cap analogs are able to generate immunogenic protein for a long time span and create a dynamic positive impact on the designing of personalized medicines. There has been growing worldwide concern to prevent infectious diseases due to the recent outbreak of SARS-CoV-2, Dengue, and Ebola virus. The mRNA vaccination provides a powerful weapon to control infectious diseases. [57] Given the recent FDA approval of mRNA vaccine for SARS-CoV-2, the design and development of novel trinucleotide cap analogs are warranted in the area of mRNA vaccines. One of the biggest challenges in this area is the isolation of 100 % capped mRNA from uncapped mRNA, wherein the uncapped mRNA negatively influences the overall translational efficiency.

[13h] ] The development of a practical and feasible method is demanded to isolate the capped mRNA from uncapped mRNA. In addition to the mRNA vaccine, the exploration of mRNA as a vehicle for antigen delivery to dendritic cells is one of the thrust areas of research in view of creating a novel immunotherapeutic approach. [58] Furthermore, considerable efforts should be invested to explore cap analog as a potential inhibitor to target eIF4E protein [59] and antiviral target. [60] The preliminary encouraging results of cell penetrating peptide as a carrier for the internalization of cap analogs open up a new avenue to design novel anticancer drug. [42 Many new therapeutic applications are emerging in the area of mRNA cap analogs and in the next two to five years will see cap analog as a candidate for clinical trials for immunotherapy and anticancer therapy.

Proc. Natl. Acad. Sci

Charles; f) M. Shanmugasundaram

Xibao Shengwuxue Zazhi

Version of record online

Trinucleotide cap analogs outperform dinucleotide cap analogs in terms of capping efficiency and translational properties. mRNA vaccines contain the presence of mRNA cap 1 structure. mRNA cap analogs can be used in the area of anticancer immunization, protein production, and gene therapy.