What this is
- This research develops a novel method for purifying capped using hydrophobic photocleavable cap analogs.
- The PureCap method achieves 100% , overcoming limitations of existing techniques.
- It allows for the effective separation of capped from uncapped , which is vital for therapeutics.
Essence
- The PureCap method enables the purification of fully capped with 100% efficiency while eliminating immunogenic uncapped . Cap-2 produced using this method demonstrates up to 4× higher translation activity compared to Cap-1 .
Key takeaways
- PureCap technology achieves 100% for various lengths, allowing for effective purification without enzymatic treatment.
- Cap-2 prepared using PureCap shows a 3- to 4× increase in translation activity in cultured cells and animals compared to Cap-1 from standard methods.
- The method reduces immunogenicity by removing both uncapped 5′ ppp-RNA and double-stranded RNA contaminants during purification.
Caveats
- The incorporation efficiency of tetranucleotide cap analogs remains low, at about 25%, which may limit the method's applicability.
- Scalability of the PureCap purification process using RP-HPLC could pose challenges for commercial production.
Definitions
- mRNA: Messenger RNA, a type of RNA that conveys genetic information from DNA to the ribosome for protein synthesis.
- Capping efficiency: The percentage of mRNA molecules that have a proper 5' cap structure, which is crucial for stability and translation.
Simplified
Introduction
In chemical or semi-chemical methods of capped RNA synthesis14–17, the length of the RNA strand is limited to about 150 bases. For synthesizing longer capped mRNA with high efficiency, RNA polymerase of phage origin is mainly used, especially T7 RNA polymerase (T7 RNAP). The current capping methods include enzymatic and co-transcriptional capping5. In the enzymatic capping, transcribed mRNA is treated with a capping enzyme5,18, often followed by methyltransferase treatment to convert the Cap-0 structure to the Cap-15. In co-transcriptional capping, a synthetic cap analog, i.e., m7G-containing nucleotide, is added to the in vitro transcription (IVT) reaction19,20. The first generation dinucleotide cap analog m7G(5′)ppp(5′)G had the problem of the chain-elongation starting from the 3′ hydroxyl group of the m7G. The unintended elongation results in 30–50% of the reversely capped RNA being inactive for translation20. This issue was solved by an anti-reverse cap analog [ARCA, m27,3’-OG(5’)ppp(5’)G] methylated in the 3′-hydroxyl group of m7G, ensuring the chain-elongation exclusively from the desired side21,22. Recently, new cap analogs consisting of trinucleotides and tetranucleotides have been developed23–27 to introduce the Cap-1 or Cap-2 structure into mRNA directly. Despite such progress, co-transcriptional capping still has a challenge in obtaining capped mRNA with 100% efficiency because nucleotides such as GTP compete with cap analogs for the initiation step of the transcription reaction.
The uncapped byproduct has a triphosphate (ppp) at the 5′ end, a motif found in viral mRNA, triggering an immune response via innate immune receptors such as RIG-1 and melanoma differentiation-associated protein 5 (MDA5)11,28. Moradian et al. have reported that a small amount of 5′ ppp-RNA byproduct from a standard co-transcriptional capping method induces substantial immune responses29, wherein the capping method provides around 80–90% capping efficiency30. These studies highlight the necessity to maximize capping efficiency or minimize uncapped ppp-RNA contaminants. The enzymatic treatment is a standard method to remove uncapped mRNA byproducts. For example, the 5′-terminal triphosphate of uncapped RNA will be converted to the monophosphate by RNA 5′ polyphosphatase, followed by degrading monophosphorylated RNA using XRN I, a 5′→3′ exoribonuclease requiring 5’ monophosphate24,26,27,31. However, considering mRNA susceptibility to non-specific degradation, the number of purification processes should be minimized. In addition, repeated enzymatic treatment significantly increases the production cost, hampering commercial mass preparation of mRNA. These issues motivate us to develop a non-enzymatic methodology for diminishing the uncapped mRNA immunogenicity.
Physical separation of capped mRNA from uncapped mRNA would become a simple, robust, cost-effective purification strategy. However, capped and uncapped mRNA have almost the same physicochemical properties, posing challenges for their separation.
In this study, we design hydrophobic cap analogs, namely PureCap analogs, for purifying capped mRNA. PureCap cap analogs are integrated into the mRNA 5’-terminal during the IVT reaction. Its hydrophobicity enables the purification of the capped mRNA using reversed-phase high-performance liquid chromatography (RP-HPLC) (Fig. 1b, c). Using a photodegradable 2-nitrobenzyl derivative as a hydrophobic tag enables tag removal under mild conditions by light irradiation32. This method relies on RP-HPLC, a standard purification process for nucleic acid medicine, rather than enzymatic treatment. Especially, purification of mRNA by RP-HPLC is an established method for eliminating double-stranded RNA (dsRNA) contaminants produced during the IVT, which increases the immunogenicity of mRNA and inhibits its translation33–37. Thus, the PureCap method can isolate capped mRNA from the two significant impurities, uncapped ppp-RNA and dsRNA, simultaneously providing a practical advantage in mRNA production.
Using the PureCap technology, we succeeded in footprint-free purification of the capped mRNA in a native form with 100% capping efficiency, which to the best of our knowledge, has not been achieved previously for the first time. The PureCap mRNA shows undetectable immunogenicity even without enzymatic treatment. Achieving 100% efficiency in various cap species, our approach provides an unbiased platform to study their structure-activity relationship by excluding the influence of different capping efficiency observed among cap species. Notably, we prepare fully capped Cap-2 mRNA, which shows up to three- to four-fold higher translation activity in cultured cells and animals compared with mRNA prepared by standard capping methods.

Hydrophobic cap analogs for the purification of capped mRNA using RP-HPLC (the PureCap method). Structure of the 5’ cap in eukaryotic mRNA.Capped mRNA can be isolated from uncapped impurities using RP-HPLC by the hydrophobicity of a tag attached to the cap moiety. After the isolation of the capped mRNA, light irradiation can remove the tag.Structure of cap analogs modified with a photodegradable hydrophobic tag, “the PureCaps,” and untagged control cap analogs synthesized/evaluated in this study. a b c
Results
Design and synthesis of dinucleotide PureCap analogs
The key step in synthesizing cap analogs is synthesizing the corresponding diphosphate by phosphorylation of guanosine derivatives. For the synthesis of cap analogs known as the ARCA derivatives, protocols consisting of monophosphorylation by the Yoshikawa method using phosphoryl chloride, activation of monophosphate through phosphorimidazolide formation, and diphosphate formation by reaction with alkylammonium phosphate were applied19. However, this method is complicated because it is based on the stepwise introduction of phosphate groups. It requires multiple purifications with aqueous solvents, such as ion exchange and reversed-phase chromatography. The introduction of a lipophilic moiety, such as in the PureCap analogs, is expected to decrease the solubility of the intermediates in aqueous solvents, making purification particularly difficult. Therefore, we applied a one-pot synthesis of the diphosphate directly from the nucleoside to avoid the solubility problem and simplify and shorten the synthetic processes. In the first approach, phosphoryl chloride is applied to a guanosine derivative, and the diphosphate is obtained by adding an alkylammonium phosphate salt to the resulting phosphorodichloridate intermediate40–42. The second approach is to obtain the diphosphate directly from 5′-tosylated guanosine by reacting with tetrabutylammonium pyrophosphate salt43–46. The resulting diphosphate is converted to the desired dinucleotide PureCap analog by methylation at the N7 position and subsequent condensation reaction with guanosine monophosphate imidazolide in the presence of zinc chloride.
2′-O-Nb-modified guanosine 14 as a phosphorylation precursor was synthesized from N2-isobutyrylguanosine 11 in four steps (Fig. 2a). First, N2-isobutyrylguanosine 11 was reacted with 1,3-dichloro-1,1,3,3-tetraisopropyldisiloxane (TIPDSCl2) in pyridine to afford the 3′,5′-O-protected compound in a quantitative yield47. The key Nb group was introduced by reaction with nitrobenzyl alcohol methylthioacetal 12 in the presence of N-iodosuccinimide (NIS) and TfOH in THF at −40 °C to afford the corresponding product 13 in 76% yield. The silyl protecting group was subsequently removed with tetrabutylammonium fluoride to afford the unprotected compound in 75% yield, which was then treated with 28% aqueous ammonia at 55 °C to remove the isobutyryl group to afford 2′-O-Nb modified guanosine 14 in 70% yield. In the first approach, 14 was reacted with phosphoryl chloride in trimethyl phosphate, and then tetrabutylammonium monophosphate salt was applied to afford the corresponding diphosphate 15 in one pot after hydrolysis (Supplementary Fig. 3). Although 15 could be synthesized by this method in a short step, the low yield of the diphosphate formation (13%) was a problem. Therefore, direct diphosphorylation via 5′-O-tosylated form 14 was investigated as a second approach. First, 2′-O-Nb guanosine (14) was reacted with tosyl chloride in pyridine to obtain the 5′-O-tosylated form in 60% yield, which was then reacted with tetrabutylammonium diphosphate salt to give the diphosphate 15 in 54% yield. The resulting diphosphate was reacted with methyl iodide in DMSO to afford the 7-methylguanosine diphosphate derivative in 72% yield. This compound was then converted to the corresponding phosphorimidazolide 16 by treatment with imidazole, 2,2′-dithiodipyridine, and triphenylphosphine in 89% yield. Finally, compound 17 was reacted with guanosine monophosphate phosphoroimidazolide 18 in the presence of zinc chloride to afford PureCap analog DiPure (1) in 50% yield (Fig. 2b). 2′-O-methylated analog DiPure/2’OMe (2), 3′-O-methylated analog DiPure/3’OMe (3), and N2-Nb-protected analog DiPure/N2 (4) were synthesized similarly (Supplementary Figs. 6–8).

Synthesis of PureCap analogs. Synthesis of 7-methyl 2’--Nb-guanosine 5’ diphosphate imidazolide ().Synthesis of dinucleotide PureCap analog DiPure ().Synthesis of tri/tetranucleotide PureCap analogs; TriPure_0 (), TriPure_1 (), TetraPure_2 (), and TetraPure_2/ m6A (). a 16 b 1 c 5 6 7 8 O
Design and synthesis of tri/tetranucleotide PureCap analogs
The cap structure produced by the co-transcription of dinucleotide cap analogs is Cap-0. However, in 2009, Ishikawa et al. reported that mRNAs with the Cap-1 structures could be prepared using trinucleotide cap analogs23. Sikorski et al. recently synthesized a series of trinucleotide cap analogs and compared their translation activity of mRNA containing Cap-1 with that of mRNA containing Cap-0, reporting a higher translation activity of the former in some mammalian cultured cells24. Methylation of the terminus 3′ hydroxyl of the trinucleotide cap analogs cannot introduce the Cap-2 structure into mRNA, as is known as the rationale for developing ARCA21,22,39. To this end, tetranucleotide cap analog are required to introduce a Cap-2 structure in mRNA, as reported recently by ref. 27.
We synthesized tri- or tetranucleotide PureCap analogs in this study to produce quantitatively capped mRNAs with Cap-1 and Cap-2 structures (Figs. 1c, 2c). We selected A and m6A bases as the base of the sites to which m7G links in the tri- or tetranucleotides; according to the report that the highest translation activity with Cap-1 mRNA was observed when the base in the cap structure was A or m6A24.
Sikorski et al. reported the synthesis of trinucleotide cap analogs by reacting 5′-monophosphorylated dinucleotides synthesized by a solid phase method with N7-methylguanosine diphosphate imidazolide in the presence of zinc chloride24. Senthilvelan et al. also reported the synthesis of trinucleotide cap analogs using 5′-monophosphorylated dinucleotides synthesized by a liquid phase method26. The 5′-monophosphorylated dinucleotides (19, 20) and 5′-monophosphorylated trinucleotides (21, 22) synthesized by these synthetic methods were reacted with N7-methyl-2’-O-Nb-guanosine diphosphate imidazolide 16 in the presence of zinc chloride to synthesize PureCap analogs in trinucleotide (5,6) and tetranucleotide (7,8) forms. (Fig. 2c).
Preparation and purification of mRNA using dinucleotide PureCap analogs
DiPure/N2, a cap analog with a protecting group at the N2 position of m7G, was similarly added to the IVT reaction, and the generated RNA was analyzed by RP- HPLC. As a result, it was found that capped mRNA could be isolated and purified for mRNA containing this analog using the hydrophobicity of the cap analog (Fig. 3e).
When T7 RNAP transcribes RNA, it is known that highly immunogenic dsRNA is produced as a byproduct33–37. When this dsRNA is included in the mRNA as an impurity, the translation activity of the mRNA is significantly impaired due to the triggering of a cellular immune response24,33,35. Purification by RP-HPLC or removal using cellulose powder is known for removing dsRNA impurity24,33,35. In our preliminary study, we also noticed that purifying capped RNA by RP-HPLC increased the translation activity in HeLa cells by about 50–100-fold. Detection experiments using anti-dsRNA J2 antibody showed that this increase in activity was due to the removal of dsRNA by RP-HPLC purification. Using RNA transcripts with no cap analog added, we examined the elution time of the dsRNA impurity by RP-HPLC and found that it eluted about 2–3 min later than the main product, 650-nt mRNA, under our analytical conditions (Supplementary Fig. 15). This position coincides with the elution position of mRNA containing PureCap analogs. Therefore, mRNA containing PureCap had to be purified once by HPLC preparative purification of capped mRNA using its hydrophobicity, followed by deprotection by light irradiation. Then a second HPLC purification was added for complete dsRNA removal. Control mRNA samples were also purified by HPLC twice to match the number of cycles in the HPLC purification process.
Capped Nluc mRNAs using DiPure/3’OMe and DiPure/2’OMe were also prepared (Supplementary Fig.). In this case, we examined the transcription reaction conditions. We found that the capped mRNA yield after RP-HPLC isolation increased if the GTP concentration was not reduced in the reaction and kept at 2 mM, the same as other NTPs, although the incorporation rate of the PureCap analog decreased. Therefore, in subsequent experiments, the GTP concentration was not reduced, and the desired pure-capped mRNA was isolated from the resulting RNA transcript mixture by RP-HPLC. 16
The purity of the resulting capped mRNAs was confirmed to be nearly identical by dPAGE (Fig. 3h). A detection assay for dsRNA contamination using the anti-dsRNA antibody confirmed that dsRNA was similarly removed from the capped mRNAs (Fig. 3i). These highly purified capped mRNAs were used for subsequent immunogenicity and translation activity evaluations.

IVT and purification of mRNA by RP-HPLC using dinucleotide PureCap analogs. Nluc mRNA (650-nt) was transcribed using T7 RNAP in the absence or presence of a cap analog.Representative dPAGE analysis of the transcription reaction.‒RP-HPLC analysis of the RNA transcript. A cap analog added to the reaction was indicated in the figure. The transcript was analyzed as a crude mixture (, black line) or after being purified by preparative HPLC (, blue line). Purified RNA was analyzed after deprotection by irradiating 365 nm light (, green line). The elution time (min) of the peak was noted nearby. A ratio (%) of capped mRNA calculated based on the peak area was listed in parentheses after the elution time. The elution time in the experiment shown indiffered from others (‒) because HPLC conditions were not completely identical.‒Analysis of purified mRNA used to measure the biological activities. The mRNAs were named after the cap analog used. Cap(−): mRNA prepared without cap analog. AnP: mRNA treated with Antarctic phosphatase. DiPure (mixture): mRNA transcribed in the presence of DiPure but without isolation of capped mRNA. (,) Analysis of the 5′ end of purified mRNA cleaved with DNAzyme 10–23 to assess its capping state.dPAGE analysis of the cleaved 5’ RNA fragments.MALDI-TOF MS analysis of the 5′ RNA fragments. For these observed peaks, the calculated and measured MS agreed well, shown in Supplementary Table.dPAGE analysis of purified mRNAs (50 ng each). Lane 1, Cap(−); 2, ARCA; 3, ARCA/AnP; 4, DiPure; 5, DiPure/N2; 6, DiPure/3’OMe; 7, DiPure/2’OMe; 8, DiPure (mixture). () Removal of dsRNA from purified mRNA (250 ng each) was confirmed by dot-blot assay using anti-dsRNA J2 antibody. As a positive control of the assay, the dsRNA sample was spotted in spots 1 to 5 (0, 0.050, 0.50, 5.0, and 50 ng, respectively). Analyzed mRNAs were as follows; spot 6, mRNA before HPLC purification prepared with no cap analog; 7, Cap(−); 8, ARCA; 9, ARCA/AnP; 10, DiPure; 11, DiPure/N2; 12, DiPure/3’OMe; 13, DiPure/2’OMe; 14, DiPure (mixture).‒Each experiment was repeated independently at least three times to obtain similar results. Source Data are provided with this paper. a b e i ii iii e b d f i f g f g h i a g 1
Preparation and purification of mRNA using tri/tetranucleotide PureCap analogs

Tri/tetranucleotide PureCap analogs are incorporated by T7 RNAP in a model system. Three types of promoter sequences were tested using 60-bp dsDNA. The estimated position of cap analogs when they were incorporated into RNA was depicted in the figure. Major or minor species were shown in blue or gray color, respectively.,dPAGE analysis of the reaction. An aliquot was taken from the reaction mixture and analyzed (upper panel in,). Or they were analyzed after being cleaved by DNAzyme 10–23 to align the heterogeneous 3′ ends of the RNA transcripts (lower panel in).Effect of the promoter sequence on the incorporation of the cap analogs. The concentrations of NTPs and cap analogs were 2 mM.Analysis of IVT on the template with Type III ϕ6.5 promoter. NTP other than GTP and cap analog concentrations were set at 2 mM, and GTP concentrations were set at either 0.5 or 2 mM. Each experiment was repeated independently at least three times to obtain similar results. Source Data are provided with this paper. a b c b c b b c

Nluc mRNA (650-nt) preparation using tri/tetranucleotide PureCap analogs. –RP-HPLC analysis of IVT-RNAs. PureCap analogs TriPure_0 (), TriPure_1 (), TetraPure_2 (), and TeraPure_2/m6A () were added to the reaction. Control mRNA using cap analogs Tri_1 () or Tetra_2 () was transcribed from a DNA template containing the A-inserted promoter. The transcript was analyzed as a crude mixture (, black line) or after being purified by preparative HPLC (, blue line). Purified RNA was analyzed after deprotection by irradiating 365-nm light (, green line). The elution time (min) of the peak was noted nearby. A ratio (%) of capped mRNA calculated based on the peak area was listed in parentheses after the elution time.‒Analysis of purified mRNA used to measure the biological activities. The mRNAs were named after the cap analog used. Cap(−) means no cap analog was added. AnP means it was further treated with Antarctic phosphatase.dPAGE analysis of the 5′ end of purified mRNA cleaved with DNAzyme 10–23 to assess its capping state. The capped ratio was calculated based on the band intensities of the capped and uncapped RNA fragments.dPAGE analysis of purified mRNAs (25 ng each). Lane 1, Cap(−); 2, TriPure_0; 3, TriPure_1; 4, TetraPure_2; 5, TetraPure_2/m6A; 6, Tri_1; 7, Tri_1/AnP; 8, Tetra_2; 9, Tetra_2/AnP.Removal of dsRNA from purified mRNA was confirmed by dot-blot assay using anti-dsRNA J2 antibody. 250 ng mRNAs were spotted. As a positive control of the assay, the dsRNA sample was dotted in spots 1 to 5 (0, 0.050, 0.50, 5.0, and 50 ng, respectively). Analyzed mRNAs were as follows; spot 6, mRNA before HPLC purification prepared with no cap analog; 7, Cap(−); 8, TriPure_0; 9, TriPure_1; 10, TetraPure_2; 11, TetraPure_2/m6A; 12, Tri_1; 13, Tri_1/AnP; 14, Tetra_2; 15, Tetra_2/AnP.‒Each experiment was repeated independently at least three times to obtain similar results. Source Data were provided with this paper. a f a b c d e f i ii iii g i g h h a i
Immunogenicity of mRNAs prepared using PureCap analogs
Mammalian cells have an immune system as a defense mechanism against infection by foreign organisms such as viruses11,28. Externally introduced RNA is recognized by pattern recognition receptors such as Toll-like receptor 3 (TLR3), TLR7, TLR8, and RIG-I, triggering an immune response in host cells28. If the prepared mRNA contains uncapped 5′ ppp-RNA and/or dsRNA impurity, the defense mechanism against antiviral infection will be activated11,28. When they are sensed by RIG-1, a cellular-type immune response occurs, activating protein kinase R and inhibiting intracellular translation. 5′ ppp-RNA is also recognized by the antiviral proteins Interferon-induced protein with tetratricopeptide repeats 1 (IFIT1) and IFIT554. Therefore, removing as much uncapped mRNA as possible when preparing capped mRNAs is essential, which may cause unwanted immune responses.
These experiments demonstrate the effect of removing 5’-triphosphate from uncapped mRNA, which is included as an impurity, on lowering the immunostimulatory effect of prepared mRNA. At the same time, they show the effectiveness of the PureCap method, which can remove both 5′-ppp-RNA and dsRNA impurities in the RP-HPLC purification process.

Evaluation of the immunostimulatory effect of capped mRNAs using NF-κB reporter (Luc)‒HEK293 cells. 650-nt Nluc mRNA prepared/ purified using dinucleotide cap analogs () or tri/tetranucleotide cap analogs () was transfected into the cells using Lipofectamine MessengerMAX reagent. Poly I: C was used as the positive control of this assay. The cap analog used for mRNA preparation is indicated in the graph. AnP means the mRNA was further treated with Antarctic phosphatase to remove its 5′ phosphate. ARCA-capped RNA before the HPLC purification was also analyzed (ARCA, crude). The expression level was normalized to the mock-transfected control sample (No RNA). Data were mean ± s. e. m. (standard error of the mean) for biological replicates (= 3 or 4). The statistically significant differences of Nluc mRNA-transfected samples from the mock-transfected sample in the one-way ANOVA test followed by the Dunnett test were marked as follows: ns,> 0.05; *,< 0.05; **,< 0.01; ***,< 0.001. Each experiment was repeated independently at least twice to obtain similar results. Source Data are provided with this paper. a b n p p p p
Translation activity measurement of Nluc mRNAs prepared by the PureCap method
Next, we checked the effect of increasing capping efficiency on mRNA translation activity (Fig. 7b). We prepared a pair of mRNAs differing only in capping efficiency, with almost 100% in one and lower in the other. Firstly, mRNA transcribed using DiPure cap analog (1) was divided into two groups. From one, capped mRNA was purified using RP-HPLC followed by photo-irradiation. This process provided mRNA with over 99% capping efficiency, which is denoted as DiPure. The other group of mRNA was photo-irradiated before RP-HPLC purification. This method provides a single peak in RP-HPLC containing the mixture of uncapped and capped mRNA, as photo-irradiation removes a hydrophobic tag from capped mRNA. The resulting mRNA mixture showed a capping efficiency of 57%, denoted as DiPure (mixture). DiPure and DiPure (mixture) originated from the same mRNA stock, possessed the same cap structure, and similarly received HPLC purification but differed only in capping efficiency. After the introduction to HeLa cells, DiPure (mixture) mRNA showed approximately 60% translation efficiency of that obtained by DiPure mRNA with 100% capping efficiency. The difference in capping efficiency may explain the difference in translation activity in this result.
To investigate the effect of the methyl group introduced into the 2′ or 3′ hydroxyl of m7G on the translation activity, mRNAs prepared with three PureCap analogs, DiPure, DiPure/2′OMe, and DiPure/3′OMe were compared (Fig. 7c, d, Table 1, and Supplementary Fig. 19a). In cultured HeLa and JAWS II cells, O-methylation at the 2′ positions slightly reduced its translation activity. Still, no apparent change in the activity was observed when one methyl group was introduced into the cap structure58. These experimental results are consistent with in silico docking simulations showing that the methyl group introduced at the 3′-O position of m7G of dinucleotide cap analog would not interfere with the interaction of the cap-binding protein V39 of cap-specific mRNA 2′-O-methyltransferase or the eukaryotic translation initiation factor eIF4E22. In the end, we conclude that DiPure is the best of the three analogs because it is the easiest to synthesize, has the best incorporation by T7 RNAP, and exhibits high mRNA translation activity.
The translation activity of Nluc mRNAs with Cap-0, Cap-1, and Cap-2 cap structures was compared using cultured HeLa and JAWS II cells (Fig. 7e, f and Table 1). In these experiments, unlike previous reports24, there was no apparent difference in activity between mRNAs with Cap-0 and Cap-1 in both HeLa and JAWS II cells. On the other hand, the translation activity of Cap-2-mRNAs introduced using the tetranucleotide PureCap analogs TetraPure_2 and TetraPure_2/m6A were higher than that of mRNAs with Cap-1 and Cap-027. The enhanced translation activity of Cap-2 mRNA was more pronounced in JAWS II cells than in HeLa cells and was up to 4.2-fold more active than Cap-1 mRNA prepared with Tri-1 cap analog. The results of our experiment, in which the translation activity of mRNA containing the Cap-2 structure is higher than that having the Cap-1 structure in JAWS II cells, differ from the results of a study reported by ref. 27. Still, the reason for this difference is unclear at this point. For control mRNAs prepared from a DNA template containing the A-inserted promoter, the translation activity of Cap-1 mRNA prepared using the Tri_1 trinucleotide cap, which had a high capping efficiency of 90%25, was comparable to that of mRNA prepared using the PureCap analog TriPure_1, both in HeLa and JAWS II cells. The activity of Cap-2 mRNA prepared with the tetranucleotide cap Tetra_2, in which only about 50% of the cap was introduced, was only about 1/3 of that of Cap-2 mRNA prepared with TetraPure_2. This result reflects the low capping efficiency of tetranucleotide cap analogs and indicates the PureCap method’s usefulness in introducing Cap-2 structures using tetranucleotide cap analogs with low incorporation efficiency. As for Cap-2 mRNA containing m6A prepared using TetraPure_2/m6A, a slight increase when compared mRNA with TetraPure_2 in translation activity was observed in JAWS II cells.
We also compared the translation activity of mRNAs with this Cap-0, 1, and 2 structures in a cell-free translation system based on HeLa S3 cells59. In this experiment, Cap-2-mRNAs also showed higher translation activity than Cap-0/1 mRNA, especially the Cap-2 with m6A (Supplementary Fig. 19b).

Translation activities of Nluc mRNAs in cultured mammalian cells. Capped mRNAs were introduced into HeLa cells (‒,) or JAWS II cells (,) using Lipofectamine MessengerMAX reagent. Cells were cultured for the indicated time, lysed, and the Nluc expression was measured. The cap analog used for mRNA preparation is indicated in the graph as the name of the mRNA. AnP means the mRNA was further treated with Antarctic phosphatase to remove its 5’ phosphate. Data were mean ± s. e. m. for biological replicates (= 4).,The statistically significant differences for the capped RNA samples in one-way ANOVA followed by Tukey’s test were marked as follows. ns, >0.05; *< 0.05; **< 0.01; ***< 0.001.‒The statistically significant differenesor the capped RNA samples from “ARCA” (,) or “Tri_1” (,) in one-way ANOVA followeby Dunnett’s test were marked as follows. ns, >0.05; *< 0.05; **< 0.01; ***< 0.001.‒Each experiment was repeated independently twice (,) or at least three times (‒) to obtain similar results. Source Data are provided with this paper. a c e d f a b c f c f c d e f d a f a b c f n p p p p p p

Translation activity of 650-nt Nluc mRNA in mice. The expression level of Nluc was evaluated four hours after intravenous injection of mRNA lipid nanoparticles (LNPs) into mice. The cap analog used for mRNA preparation is indicated in the Figure. AnP means the mRNA was further treated with Antarctic phosphatase. () In vivo images of mice. () Quantifying Nluc expression levels using homogenates of the liver () and spleen (). Data are mean ± s. e. m. for biological replicates (= 8). The statistically significant differences in one-way ANOVA followed by Tukey’s test were marked as follows: ns,> 0.05 (not significant); *,< 0.05; **,< 0.01; ***,< 0.001, ****,< 0.0001. (‒) Each experiment was repeated independently at least twice to obtain similar results. Source Data are provided with this paper. a b, c b c a c n p p p p p
| Name of mRNAa | Capping efficiency (%)b | Normalized Nluc expressionc | |||
|---|---|---|---|---|---|
| HeLa cells | JAWS II cells | Mice liver | Mice spleen | ||
| Dinucleotide Cap analogs (mG-ppp-G)7 | |||||
| ARCA | 56 | 1.00 ± 0.02 | 1.00 ± 0.03 | n.d. | n.d. |
| ARCA/ AnP | 56 | 0.97 ± 0.03 | 1.13 ± 0.03 | n.d. | n.d. |
| DiPure | >99 | 2.36 ± 0.05 | 3.51 ± 0.09 | n.d. | n.d. |
| DiPure/3′OMe | >99 | 2.27 ± 0.07 | 3.11 ± 0.21 | n.d. | n.d. |
| DiPure/2′OMe | >99 | 2.01 ± 0.09 | 2.58 ± 0.08 | n.d. | n.d. |
| Tri/ tetranucleotide cap analogs mG-ppp-AG(G)7 | |||||
| Tri_1 | 87 | 1.00 ± 0.04 | 1.00 ± 0.05 | 1.00 ± 0.15 | 1.00 ± 0.13 |
| Tri_1/ AnP | 87 | 1.07 ± 0.03 | 1.51 ± 0.06 | 1.47 ± 0.23 | 0.61 ± 0.08 |
| Tetra_2 | 52 | 0.52 ± 0.02 | 0.70 ± 0.04 | n.d. | n.d. |
| Tetra_2/AnP | 52 | 0.51 ± 0.02 | 0.53 ± 0.02 | n.d. | n.d. |
| TriPure_0 | >99 | 1.12 ± 0.06 | 1.04 ± 0.08 | n.d. | n.d. |
| TriPure_1 | >99 | 1.31 ± 0.05 | 1.07 ± 0.05 | 0.80 ± 0.08 | 0.90 ± 0.07 |
| TetraPure_2 | 98 | 1.91 ± 0.06 | 2.99 ± 0.19 | 4.86 ± 0.61 | 2.48 ± 0.29 |
| TetraPure_2/m6A | 95 | 1.98 ± 0.07 | 3.60 ± 0.19 | 4.07 ± 0.48 | 3.58 ± 0.63 |
Preparation/purification of longer mRNA by the PureCap method

Preparation and purification of capped 4247-nt mRNA encoding spike protein of SARS-CoV-2 using dinucleotide PureCap analogs. Chemical structure of the cap analogs with higher hydrophobicity (‒) compared to the original dinucleotide analog ().Denaturing agarose gel electrophoresis of the IVT reactions in the presence of these cap analogs.RP-HPLC analysis of IVT-RNAs with or without cap analogs. Elution time (min) and the ratio of capped mRNA were noted near the peak.Analysis of 5′ terminus of mRNA after RP-HPLC purification. After being cleaved by DNAzyme 10–23, it was analyzed by 15% dPAGE.‒Each experiment was repeated independently at least twice to obtain similar results. Source Data are provided with this paper. a 23 25 1 b c d b d
Discussion
Various cap analogs with different structural features have been developed to prepare capped mRNAs by IVT19. First to mention are the dinucleotide cap analogs ARCAs21,22,39,58,60, which were designed to control the direction of introduction of the dinucleotide cap analog m7G(5′)ppp(5′)G. It was reported that methylation of the 2′ or 3′ hydroxyl group completely controlled the direction of strand elongation, resulting in an approximately 2–3-fold increase in mRNA translation activity in cultured cells58,60. Following the successful development of ARCA, many modified analogs for dinucleotide caps have been developed. Many examples of modification of the phosphate moieties61,62. Extensions of triphosphate linkage to tetra- and pentaphosphates were reported, and improved translation activity of capped mRNA prepared using these analogs was reported39. The development of analogs in which bridging or non-bridging oxygen atoms were replaced by S, Se, CH2, and BH3 was also reported61–63. Synthesis of cap analogs in which a non-phosphate linkage was introduced into the phosphate moiety by the click reaction has been reported64,65. These modifications in the linkage often increase the translation activity of the mRNA, mainly because they increase its resistance to degradation by the scavenger mRNA-decapping enzyme DcpS. Synthesis of functional cap analogs that can create labeled translatable mRNA by introducing fluorescent groups or biotin into the 2′ and 3′ hydroxyl groups of m7G was also reported63,66,67. The introduction of a benzyl group at the N2 position of m7G or the substitution of a 7-methyl group with an arylmethyl group was reported to enhance the translation activity of mRNA in human cultured cells58,68,69. Some cap analogs have been developed for on/off control of the mRNA translation31,70–72. For example, Ogasawara et al. reported the synthesis of cap analogs that contained photo-responsible groups such as phyenylazo or styryl group or at the N2 or C8 position of m7G and that their cis/trans conformation enabled photo-switchable translation control70–72. As the biological significance of the Cap-1 and Cap-2 structures becomes clearer, tri- and tetranucleotide cap analogs have also been developed to introduce them directly into mRNA23–27.
Compared to the development of the cap structures mentioned above, much less effort has been devoted to improving capping efficiency25,67. Currently, a method to improve the incorporation efficiency of trinucleotide cap analogs exploiting their complementarity with the template is widely used25. Although the capping efficiency with this method is reported to be 80–90%30, further improvement of the capping efficiency will be necessary because of the substantial amount of contaminating 5′ ppp-RNA that induces strong innate immunity29. As is well known, it is important to remove immunostimulatory dsRNA byproducts produced by T7 RNAP during transcription33,34. We undertook the present study to remove these byproducts for the therapeutic use of mRNA. The significance of this study lies in designing cap analogs for increasing capping efficiency, while previous studies derivatized numerous cap analogs for improving cap functionalities as mentioned above; the PureCap method we are developing here is unprecedented in that it removes 5′ ppp-RNA by chromatography. Indeed, we succeeded in completely suppressing innate immune responses without enzymatic treatment, such as Antarctic phosphatase (Fig. 6). Further notably, the hydrophobic tag introduced for purification can be removed simply by light irradiation, leaving no unnatural modification in the capped mRNA.
PureCap technology also provides a reliable platform to study structure-activity and the relationship of cap analogs. In co-transcriptional capping, different cap structures resulted in different capping efficiency (Fig. 5g), causing bias in their comparison. Indeed, our PureCap method achieved almost 100% capping efficiency in various capping structures, including Cap-0, Cap-1, and Cap-2, with or without modification in 2′ or 3′ hydroxyl of m7G and N6 position of the first transcribed A (m6A). Especially, quantitative capping was obtained in preparing Cap-2 mRNA, which shows low capping efficiency in the conventional method (Fig. 5g). Cap-2 mRNA thus prepared showed up to three- to four-fold higher translation activity in cultured cells and animals than mRNA prepared by standard capping methods without inducing innate immune responses (Figs. 6–8). Intriguingly, Cap-2 mRNA improved protein translation efficiency by approximately threefold in the liver and spleen, two major target organs of mRNA therapeutics. Numerous studies and clinical developments target the liver for treating single gene disorders providing therapeutic proteins systemically73,74, and the spleen for cancer vaccines75, immunotolerance vaccines76, and chimeric antigen receptor T cell potentiation77. The Cap-2 mRNA technology is capable of potentiating these therapeutics.
One issue to be resolved in the future is the low incorporation efficiency of the tetranucleotide cap analogs to give the Cap-2 structure in mRNA, which was about 25% of the total. Currently, wild-type T7 RNA polymerase is used, but in the future, we would like to improve this, for example, with modification of the polymerase78. Scalability issues in the RP-HPLC purification process may be a concern. We will be able to learn from the process of the commercial-scale manufacturing of therapeutic oligonucleotides that have recently been launched79. Due to their structural complexity, these oligonucleotides are purified by HPLC even on a commercial scale80–82. We believe that mRNA purification above the gram scale using RP-HPLC is feasible using a system similar to that used for oligonucleotide therapeutics.
In summary, we synthesized quantitatively capped mRNAs by the PureCap method using synthesized hydrophobic cap analogs. To the best of our knowledge, our system provides a footprint-free purification method of capped mRNA with 100% capping efficiency for the first time. Notably, increasing the capping efficiency of mRNA resulted in increased translation activity and reduced immunostimulation. Practically, this technology requires only RP-HPLC for purifying capped mRNA, avoiding repeated enzymatic treatment commonly performed for reducing mRNA immunogenicity. Furthermore, this approach is versatile to apply to mRNA with different lengths (650 nt and 4247 nt). Achieving 100% capping efficiency in various cap species, our approach provides an unbiased platform to study their structure-activity relationship by excluding the influence of different capping efficiency observed among cap species. In this study, we prepared fully capped Cap-2 mRNA, which shows low capping efficiency in conventional methods. Ultimately, Cap-2 mRNA thus prepared showed up to three- to four-fold higher translation activity in cultured cells and animals than mRNA prepared by standard capping methods without inducing innate immune responses. Worth notably, Cap-2 mRNA improved protein translation efficiency in the liver and spleen, two major target organs of mRNA therapeutics, providing an excellent tool for future mRNA-based medicines.
Methods
Ethical statement
The animal experiments were conducted under the approval of the animal care and use committees at the Kyoto Prefectural University of Medicine and the Innovation Center of NanoMedicine (iCONM) with all relevant ethical regulations.
Preparation of Nluc mRNA by IVT and its purification by RP-HPLC
The sequence of 650-nt Nluc mRNA is described in the Supplementary Information file. Template DNA for IVT was prepared by PCR reaction using pNL1.1 TK vector (Promega) as a template. The sequence of the PCR primers was as follows: Forward (Type III ϕ6.5 promoter-containing), 5′-CCCGGATCCTAATACGACTCACTATAGGCGCATATTAAGGTGACGCGT-3′; Reverse, 5′ (T)30-CTAGAATTACGCCAGAATGCG-3′. To prepare DNA containing different T7 promoters as Type II ϕ2.5 or the one named “A-inserted,” a different forward primer was used as described below: For Type II ϕ2.5: 5′- CCCGGATCCTAATACGACTCACTATTAGGCGCATATTAAGGTGACGCGT-3′: For “A-inserted”; 5′-CCCGGATCCTAATACGACTCACTATAAGGCGCATATTAAGGTGACGCGT-3′. The PCR mixture consisted of 0.5 μM primers, 1 ng/μL pNL1.1 TK vector, 0.2 mM dNTPs, 1.5 mM MgSO4, 1×PCR Buffer for KOD -Plus- Neo, 0.02 units/μL KOD -Plus- Neo (Toyobo). The mixture was subjected to the thermal cycling reaction as follows: 94 °C, 2 min → (98 °C, 10 s → 55 °C, 30 s →68 °C, 90 s) × 25 cycles → 68 °C, 5 min. The reaction was analyzed by agarose gel electrophoresis, and dsDNA was recovered by alcohol precipitation after TE-saturated phenol/ chloroform (1:1) extraction. A typical transcription reaction was carried out in a reaction mixture containing 15 ng/μL DNA (PCR product), 2 mM NTPs, 2 mM cap analog, 40 mM Tris-HCl (pH 8.0), 8 mM MgCl2, 2 mM spermidine, 5 mM DTT, 0.002 U/μL inorganic (yeast) pyrophosphatase (New England Biolabs), 9.4 ng/μL T7 RNA polymerase. For the control cap analogs, ARCA was purchased from Jena Bioscience, while Tri_1 and Tetra_2 were synthesized in-house. After the mixture was incubated at 37 °C for 2 h, DNase I (Takara bio) was added to the mixture at a final concentration of 0.1 units/μL, and it was further incubated at 37 °C for 15 min. After the mixture was extracted with TE-saturated phenol/chloroform and chloroform, RNA was recovered by alcohol precipitation. Progress of the IVT reaction or mRNA purity after being purified was confirmed by denaturing PAGE analysis [5(w/v) % acrylamide (ratio of acrylamide and bis-acrylamide was 19:1), 7.5 M urea, 1× Tris-borate-EDTA (TBE) buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA, pH 8.3)]. The gel was stained with SYBR Green II and visualized on the ChemiDoc Touch MP imaging system (Bio-Rad). Nluc mRNA was analyzed and purified by RP-HPLC using YMC-TriartBio C4 column (250 × 4.6 mm I.D., S-5 μm, 12 nm) on Hitachi Chromaster (pump, 5110; detector 5430) or LaChrom (pump, L-2130; detector L-2455) HPLC system with Solution_A (50 mM triethylammonium acetate (TEAA, pH 7.0) containing 5% acetonitrile) and Solution_B (acetonitrile) at a flow rate of 1 mL/min. The content of Solution_B was raised from 0 to 20% over 20 min. The column temperature was maintained at 50 °C. After purification of mRNA by RP-HPLC, RNA was recovered by alcohol precipitation from the eluate in the presence of NaOAc (pH 5.2) and 2-propanol. RNA concentration was determined by measuring absorbance at 260 nm on NanoDrop 2000 spectrophotometer (Thermo). Nb-protection of the capped mRNAs prepared using PureCap analogs was removed by 365-nm light irradiation to the mRNA. mRNA solution was transferred into a well of the 96-well multi-well plate at a volume of 50 μL/well. 365-nm light was irradiated to the RNA at 4 mW/cm2 for 10 min at room temperature using MAX-350 compact xenon light source (Asahi Spectra). After light irradiation, the progress of the deprotection reaction was confirmed by the change in the elution time in RP-HPLC analysis. The deprotected capped mRNA was purified by RP-HPLC under the same conditions as above to remove residual dsRNA contamination. RP-HPLC purified control mRNAs were dephosphorylated using Antarctic phosphatase (AnP, New England Biolabs) in a reaction mixture containing 250 ng/μL Nluc mRNA, 50 mM Bis-Tris-propane-HCl, 1 mM MgCl2, 0.1 mM ZnCl2 (pH 6), 0.25 U/μL AnP. After being incubated at 37 °C for 30 min, the mixture was extracted with TE-saturated phenol/ chloroform (1:1) and chloroform. RNA was then recovered by alcohol precipitation.
Dot-blot analysis to detect dsRNA byproduct
mRNA samples were spotted onto the Amersham Hybond-N+ membrane and air-dried. After being blocked in TBT-T [50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.05%(v/v) Tween-20] containing 5% (w/v) skim milk (Wako) for 1 h at room temperature, the blot was incubated with anti-dsRNA clone J2 (Sigma-Aldrich, MABE1134-100UL or Jena Bioscience, RNT-SCI-10010200) for 1 h at room temperature, which 1000-fold diluted with TBS-T containing 0.5% skim milk. After being washed with TBS-T, the blot was further incubated with anti-mouse IgG-HRP (Sigma-Aldrich, A9044) for 1 h at room temperature, diluted by 5000-fold with TBS-T containing 0.5% skim milk. After washing with TBS-T, the blot was incubated with SuperSignal West Femto Maximum Sensitivity Substrate (Thermo) and visualized on a ChemiDoc Touch MP imaging system (Bio-Rad). siRNA Ladder Marker (Takara, cat# 3430) was used as a positive control of dsRNA.
RNA cleavage using a DNAzyme 10–23 to analyze the 5’ end of mRNA
DNAzyme 10–23 was designed according to a previous report49. The sequences used in this study are listed as follows (catalytic domain underlined): for Nluc mRNA, 5′-TTCGAGGCCAGGCTAGCTACAACGAACGCGTCACC-3′; for 34/35-nt short RNA, 5′-TTGTAGTCCAGGCTAGCTACAACGACGGATATATCTCCT-3′; for spike protein mRNA, 5′-TCTGTGGGGAGGCTAGCTACAACGACAGAAGAATACTAG-3′. Target RNA (0.5 μM) and DNA (1 μM) were incubated in a buffer composed of 50 mM Tris-HCl (pH 8.0) and 50 mM MgCl2 at 37 °C for 1 h. The mixture was then alcohol-precipitated. 1.57 μg RNA for 650-nt Nluc mRNA was digested in a 15 μL reaction and analyzed by 15(w/v)% denaturing PAGE containing 7.5 M urea and 1×TBE. The gel was stained with SYBR Gold and visualized on a ChemiDoc Touch MP imager. As for MALDI-TOF MS analysis of the reaction, the reaction mixture (10 μL) was twice alcohol-precipitated using ammonium acetate to remove sodium ions. It was measured in a positive mode with an UltrafleXtreme MALDI-TOF/TOF mass spectrometer (Bruker Daltonics) using 3-hydroxypicolinic acid as a matrix.
Short RNA transcription system using 60-bp DNA as a template
IVT templates containing three different T7 RNA promoter sequences were prepared by the extension reaction of 26-nt DNA oligo annealed to the 60-nt oligo using DNA polymerase. The sequences of the oligos used in this experiment were listed below (5′→3′ direction; underlined promoter sequence; Nm, 2′-O-methyl-modified).Type III ϕ6.5 containing (60-nt), 5′-CCCGGATCCTAATACGACTCACTATAGGGATCCGAAGGAGATATATCCGATGGACTACAA-3′; Type II ϕ2.5 containing (60-nt), 5′-CCGGATCCTAATACGACTCACTATTAGGGATCCGAAGGAGATATATCCGATGGACTACAA-3′; A-inserted containing (60-nt), 5′-CCGGATCCTAATACGACTCACTATAAGGGATCCGAAGGAGATATATCCGATGGACTACAA-3′; Reverse strand for all (26-nt): 5′ Um_Um_GTAGTCCATCGGATATATCTCCTT-3′. 60-bp dsDNA IVT templates were prepared as follows. A reaction mixture composed of 2 μM DNA oligos (60-nt and 26-nt), 0.2 mM dNTPs, 1×PrimeSTAR HS Buffer, 0.025 U/μL PrimeSTAR HS DNA polymerase was incubated at 95 °C for 2 min, 55 °C for 1 min, and 72 °C for 60 min. It was extracted with TE-saturated phenol/chloroform (1:1) chloroform, and the DNA was alcohol-precipitated. Transcription of 34/35-nt RNA was carried out in a reaction mixture containing 15 ng/μL 60-bp dsDNA, 2 mM NTPs, 2 mM PureCap analog, 40 mM Tris-HCl (pH 8.0), 8 mM MgCl2, 2 mM spermidine, 5 mM DTT, 9.4 ng/μL T7 RNA polymerase. After incubation at 37 °C for 2 h, a portion (0.5 μL) was taken from the mixture and analyzed by 15% denaturing PAGE containing 7.5 M urea. The gel was visualized by SYBR Green II staining.
Immunogenicity measurement of mRNAs
NF-κB reporter (Luc)-HEK293 cells (BPS Bioscience, 60650) were grown in Dulbecco’s modified Eagle’s medium (DMEM; Wako) supplemented with 10% fetal bovine serum (FBS; Invitrogen) and 100 µg/ml of Hygromycin B at 37 °C under 5% CO2 atmosphere. One day before transfection, the cells were seeded in a 48-well cell culture plate at 1.0 × 105 cells/well. Just before transfection, the medium was replaced with Opti-MEM I Reduced Serum Medium (240 µL/well, Thermo). mRNA (100 ng/well) was mixed with Lipofectamine MessengerMAX (0.15 µL/well) in Opti-MEM I medium (10 µL/well) and added to cells. Three hours after the transfection, DMEM containing 20% FBS was added to the cells (250 μL per well). Twenty-four hours after transfection, the cells were lysed using Glo Lysis Buffer (50 µL/well, Promega), and the luciferase expression was measured using ONE-Glo Luciferase Assay System (Promega). The chemiluminescence was measured on the TriStar5 plate reader (Berthold). The luciferase expression was normalized by total protein concentration in the lysate measured using Pierce BCA Protein Assay Kit (Thermo).
Translation activity measurement of Nluc mRNA using cultured mammalian cells
HeLa cells (Riken Cell Bank, RCB0007) were grown in DMEM (Wako) supplemented with 10% FBS (Invitrogen) at 37 °C under a 5% CO2 atmosphere. JAWS II cells (ATCC, CRL-11904) were grown in MEM α, nucleoside (Gibco) supplemented with 10% FBS (Invitrogen) and 5 ng/mL murine GM-CSF (PeproTech) at 37 °C under a 5% CO2 atmosphere. One day before transfection, the cells were seeded in a 96-well cell culture plate, typically at 5 × 103 cells/well for HeLa cells or 1 × 104 cells/well for JAWS II cells. Just before transfection, the medium was replaced with Opti-MEM I medium (90 µL/well, Thermo). mRNA (10 ng/well) mixed with Lipofectamine MessengerMAX (0.15 µL/well) in Opti-MEM I medium (10 µL/well) was added to the cells. Three hours after the transfection, the growing medium was added to the cells (100 μL per well). At indicated time points, the cells were lysed, and the luciferase expression was measured using Nano-Glo Luciferase Assay System (Promega). The chemiluminescence was measured on the TriStar5 or TriStar2 plate reader (Berthold Technologies). Statistical analysis of the data was performed using GraphPad Prism 9 software.
Translation activity measurement of Nluc mRNA using HeLa S3 cell-based cell-free translation system
HeLa S3 cell extract was prepared based on a previous report59. HeLa S3 cells (Riken Cell Bank, RCB1525) were cultured in 1 L of S-MEM (Gibco) supplemented with 10% FBS (Corning) and 2 mM l-glutamine (Nacalai) using a spinner flask (Corning) at 37 °C under 5% CO2 atmosphere. When the cell number reached the order of the ninth power of 10, cultured cells were precipitated by centrifugation and washed three times with the washing buffer [35 mM HEPES-KOH (pH 7.5), 140 mM NaCl, and 11 mM glucose] and once with the extraction buffer [20 mM HEPES-KOH (pH 7.5), 45 mM potassium acetate, 45 mM KCl, 1.8 mM magnesium acetate, and 1 mM DTT]. Then, the cell pellet was suspended with an equal volume of the extraction buffer and still left in the pressure vessel (Parr Instrument company), kept under 500 psi of nitrogen gas for 30 min on ice. Cells were disrupted by the sudden release of nitrogen gas pressure dissolved in the cell suspension. After disruption, 1/300 volume of the high potassium buffer [20 mM HEPES-KOH (pH 7.5), 945 mM potassium acetate, 945 mM KCl, 1.8 mM magnesium acetate, and 1 mM DTT] was added to the lysate and mixed well. It was centrifuged three times at 4 °C to recover the supernatant as cell extract, being divided into aliquots and stored at −80 °C. Translation mixture composed of 0.5 μL of 30 ng/μL Nluc mRNA (650-nt), 4.35 μL of HeLa S3 cell extract, 3.25 μL of Mixure-2 [12.7 mM DTT, 5.42 mM magnesium acetate, 85.4 mM potassium acetate, 282 mM HEPES-KOH (pH 7.5)], 0.9 μL of Mixture-3 [12.6 mM ATP, 1.22 mM GTP, 200 mM creatine phosphate, 0.6 mg/mL creatine kinase, 0.08–1.2 mM 20-amino acids mixture], was incubated at 32 °C. The expression of Nluc was measured after 0.5-, 1-, 2-, 4-, and 6 h using Nano-Glo Luciferase Assay System (Promega). Statistical analysis of the data was performed using GraphPad Prism 9 software.
Translation activity measurement of Nluc mRNA using mice
For lipid nanoparticle (LNP) preparation, 595 μg/mL of mRNA was dissolved in citrate buffer (pH 3.0) and lipid mixture from D-Lin-MC3-DMA (MedChemExpress), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC, Wako), cholesterol (Sigma), and 1,2-dimyristoyl-rac-glycero-3-methoxypolyethylene glycol-2000 (DMG-PEG2000, Avanti) was dissolved in ethanol at D-Lin-MC3-DMA/DSPC/Sigma/DMG-PEG2000 molar ratio of 50/10/38.5/1.5 and total ethanol concentration of 35 mM. The citrate buffer and ethanol solution were mixed at a volume ratio of 2 to obtain a nitrogen/phosphate ratio of 5 using microfluidics (NanoAssemblr® Spark, Precision Nanosystems). The solvent was exchanged for PBS by diluting the solution 40-fold and concentrating LNP using Amicon Ultra-15–30K centrifugal units (Merck Millipore). About 2 μg of mRNA encapsulated in LNP was injected into Balb/c mouse (female, 7-week-old, Charles River Laboratories Japan, Inc. Housing conditions were maintained as follows: 12-h/12-h dark/light cycle, 23 ± 2 °C temperature and 40 to 60% humidity). For in vivo imaging, 5 μg/mouse of furimazine (Chem Shuttle) was injected from the mouse tail vein. Immediately after the injection, the mouse was observed using IVIS Spectrum SP-BFM-T1 (Perkin Elmer). The liver and spleen were excised at 4 h after the LNP injection for quantification using tissue homogenates. The organs were homogenized with Passive lysis buffer (Promega), followed by luciferase assay using Lumat3 LB9508 luminometer (Berthold Technologies) and Nano-Glo Luciferase Assay System (Promega). Statistical analysis of the data was performed using Kaleida Graph 5.0 software.
Preparation and purification of spike protein mRNA of SARS-CoV-2
Reaction conditions in the transcription of capped 4247-nt mRNA encoding spike protein of SARS-CoV-2 using cap analogs were the same as that of 650-nt Nluc mRNA. IVT template was prepared by a PCR reaction using poly-dT80-containing reverse PCR primer to introduce a poly-A80 tail to mRNA. mRNA sequence is described in the Supplementary Information file. Transcribed mRNA was analyzed and purified by RP-HPLC using YMC-TriartBio C18 column (YMC corp., 250 × 4.6 mm I.D., TA30S05-2546PTH) on Shimadzu Prominence HPLC system (pump, LC-20AD; detector, SPD-M40) with Solution_A (100 mM TEAA (pH 7.0) containing 5% acetonitrile) and Solution_B (100 mM TEAA (pH 7.0) containing 50% acetonitrile) at a flow rate of 1 mL/min. The content of Solution_B was raised from 15 to 22% over 30 min. The column temperature was maintained at 50 °C.
Statistics and reproducibility
No statistical method was used to predetermine the sample size. No data were excluded from the analyse. The experiments were not randomized. The Investigators were not blinded to allocation during experiments and outcome assessment.
Reporting summary
Further information on research design is available in thelinked to this article. Nature Portfolio Reporting Summary
Supplementary information
Supplementary Information Peer Review File Reporting Summary