Synthesis and structure elucidation of the human tRNA nucleoside mannosyl-queuosine

Queuosine (Q) is a structurally complex, non‐canonical RNA nucleoside. It is present in many eukaryotic and bacterial species, where it is part of the anticodon loop of certain tRNAs. In higher vertebrates, including humans, two further modified queuosine-derivatives exist ‐ galactosyl‐ (galQ) and mannosyl-queuosine (manQ). The function of these low abundant hypermodified RNA nucleosides remains unknown. While the structure of galQ was elucidated and confirmed by total synthesis, the reported structure of manQ still awaits confirmation. By combining total synthesis and LC-MS-co-injection experiments, together with a metabolic feeding study of labelled hexoses, we show here that the natural compound manQ isolated from mouse liver deviates from the literature-reported structure. Our data show that manQ features an α‐allyl connectivity of its sugar moiety. The yet unidentified glycosylases that attach galactose and mannose to the Q‐base therefore have a maximally different constitutional connectivity preference. Knowing the correct structure of manQ will now pave the way towards further elucidation of its biological function.

1. LC-MS-based comparison of natural manQ from mouse liver with our synthetic manQ-compounds for structure elucidation 1

.1 UHPLC-MS/MS (QQQ) based co-injection experiments for manQ structure elucidation
Supplementary Figure 1. Fragmentation pattern used for specific detection of manQ and galQ on the QQQ mass spectrometer. The fragmentation is depicted for natural manQ 4, but occurs in a similar way for the precursor ions 2, 3, 10, and 11, giving the same daughter ion 23 with m/z = 163.1 and the likely structure depicted here.
Supplementary Figure 2. UHPLC-MS/MS-based identification of the galQ-peak by co-injection. a, Digested total RNA from mouse tissue only (control); b, Digested total RNA from mouse tissue spiked with synthetic galQ as recently prepared by us 1 showed a complete signal overlap with the peak appearing at higher retention time, therefore identifying this peak as natural galQ.

HPLC-MS (Orbitrap)-based repetition of co-injection experiments for manQ structure confirmation
To double-check our results from the UHPLC-MS/MS experiments, we repeated the coinjection experiments on a second LC-MS system using a different HPLC separation column.
To take into account the lower sensitivity of the Orbitrap system in comparison to the QQQsystem, the samples, containing the digested RNA (see Supplementary Information 1 As shown in Supplementary Fig. 3, these experiments again did NOT result in the appearance of a new peak, but resulted in a full signal overlap of the natural manQ compound from mouse liver and our synthetic -allyl-manQ 4. As expected from our UHPLC-MS/MS co-injection experiments, these results thereby confirm the identity of the natural manQ compound as allyl-manQ 4. Supplementary Figure 3. Extracted ion chromatograms resulting from the co-injection of the synthetic manQ compound 4 with enzymatically digested total RNA from mouse liver analyzed by HPLC-MS: Co-injection of -allyl-manQ 4 with digested total RNA from mouse liver shows a complete signal overlap of 4 and natural manQ, thereby leading to an increased signal intensity of manQ (b) in comparison to the control sample (a) without synthetic standard. This result confirms the findings from the UHPLC-MS/MS-experiments and shows again that our synthetic α-allyl-manQ 4 is identical to the naturally occurring manQ compound. Of note, the calculated mass of α-allyl-manQ 4 and of galQ 2 is 572.2199 for [xQ + H] + . Shown here are extracted ion chromatograms with m/z = 572.2200±20, allowing for a maximal mass deviation of δ = 3.7 ppm only.
Of note, the proper HPLC-separation of the manQ-compounds on this system was more sensitive towards changes of gradient and column age than in case of the UHPLC-based separation with a Poroshell 120 SB-C8 column. In rare cases, when the column had been used for a prolonged period of time, the peaks of synthetic -homoallyl-manQ 11 and of the natural manQ compound -allyl-manQ 4 were not properly separable any more.

Investigation of manQ stability towards isomerization
To investigate whether our RNA isolation conditions could lead to isomerization of the manQisomers, we subjected compounds 3, 4, 10, and 11 to the TriReagent-isolation conditions described above. For this, TriReagent (1 mL, Sigma Aldrich) was added to a solution of each standard and vortexed for 5 min. Chloroform (200 µL) was added, the mixture was vortexed for 5 min and then kept at room temperature for 10 min. After centrifugation (12,000 g, 15 min, 4 °C), the aqueous phase was separated and isopropanol (500 µL) was added. The solution was stored at -20 °C overnight. The solvent was evaporated and the residue was dissolved in H2O (500 µL). The aqueous phase was extracted with chloroform (3 x 200 µL) and the solvent was removed by lyophilization. The residue was taken up in H2O and analyzed by UHPLC-MS/MS. The resulting chromatograms are shown in Supplementary Figure 4. None of the isomers did undergo isomerization under these conditions.
Supplementary Figure 4. Investigation of manQ stability towards isomerization under our RNA isolation conditions. Depicted are the UHPLC-MS/MS chromatograms of the stability tests. After TriReagent-treatment, all four isomers still give the expected signal at their characteristic retention time with no additional peaks of an isomerization product appearing.
Next, we checked whether our RNA digestion conditions could lead to isomerization of the manQ compounds. To this reason, a solution of the manQ nucleoside standard 3, 4, 10, or 11 in nuclease-free water (42.5 µL) was prepared. 5 µL of Nucleoside Digestion Mix Reaction Buffer (10x, New England BioLabs) and 2.5 µL of Nucleoside Digestion Mix (New England BioLabs) were added, and the mixture was incubated for 2 h at 37 °C. The sample was then 2. Metabolic feeding study to confirm mannose as the hexose-part of natural manQ Supplementary Table 1: Composition of the used media. Light RPMI medium corresponds to regular RPMI medium. For Heavy RPMI medium, a glucose free RPMI mixture was used (w/o g.: without glucose). A combination of either mannose and glucose or galactose and glucose was added to 11 mM final concentration each. One of the two hexoses was fully 13 C-labelled in each mixture.

Medium Composition
Light RPMI medium Heavy RPMI medium man/glc D-Mannose (5.00 g, 27.8 mmol, 1.00 eq) was dissolved in pyridine (104 mL). Ac2O (78.0 mL, 416 mmol, 30.0 eq) and DMAP (339 mg, 2.78 mmol, 0.10 eq) were added at 0 °C. The reaction was stirred at room temperature for 24 h, which resulted in an orange solution. After evaporation of the solvent in vacuo the residue was taken up in EtOAc (100 mL) and washed with aqueous HCl (1 M, 100 mL), H2O (2 x 100 mL) and brine (100 mL). After evaporation of the solvent in vacuo the product (10.8 g, 27.7 mmol, 99%) was obtained as yellow viscous syrup.
Rf (EtOAc/iso-hexane 1:1) = 0.76.  Analytical data in accordance with literature. (27) 3 26 (5.09 g, 11.6 mmol) was dissolved in MeOH (50 mL). A catalytic amount of sodium was added and the reaction mixture was stirred for 1.5 h at room temperature resulting in a suspension. The reaction was neutralized with AcOH, upon which the precipitate redissolved. Evaporation of the solvent gave the product (2.69 g, 11.1 mmol, 96%) as colorless foam.
Rf (MeOH/DCM 1:100) = 0.24. 12 (23 mg, 24.0 µmol was dissolved in 10% HNMe2 in THF (1 mL). The reaction was stirred for 1 h at room temperature before the solvent was removed in vacuo. The resulting yellow solid was washed with hexane to afford the product as colorless solid. The resulting Fmocdeprotected derivative (14.9 mg, 20.6 µmol, 1.00 eq) was dissolved in MeOH (0.5 mL) and 13 (14.6 mg, 20.6 µmol, 1.00 eq) and AcOH (1 µL) were added. The reaction was stirred at rt for 5 h before it was cooled to 0 °C and NaBH4 (2.02 mg, 53.6 µmol, 2.60 eq) was added. The reaction was stirred at 0 °C for 1 h. Then H2O was added and the solvent was evaporated. The residue was dissolved in MeOH (1 mL) and NaOMe (27.0 mg, 0.5 mmol) was added. The reaction mixture was stirred at room temperature for 5 h, after which LC-MS indicated deprotection of all ester type protecting groups. The solution was neutralized with DOWEX-H + -resin, then filtered and the solvent was evaporated. The residue was taken up in DCM (0.9 mL) and TFA (0.1 mL) was added at 0 °C. The reaction was stirred at 0 °C for 20 min, then saturated aqueous NaHCO3-solution was added (1 mL). The mixture was evaporated and the residue was taken up in pyridine (1.5 mL) and HF • pyridine (7.5 eq) was added in a plastic falcon. The reaction was stirred for 18 h at room temperature before TMSOMe (15.0 eq) was added at 0 °C. After stirring the reaction for 1 h at 0 °C, the solvent was evaporated in vacuo.

Synthesis of glycosyl donor 16
Supplementary Figure 9: Synthesis of glycosyl donor 16 in 4 steps starting from D-mannose (overall yield: 34 %)
Rf (EtOAc/iso-hexane 1:3) = 0.14. NOTE: Due to the inseparable anomeric mixture received, 1 H-NMR data refer to the most important features proving the formation of the glycosylation adduct from the starting materials. The anomeric mixture was separated via HPLC after completing the synthesis. 13 C-NMR-data are not assigned, but the 13 C-spectrum is depicted in section 4 below. 17 (40 mg, 42.4 µmol) was dissolved in MeCN (1.5 mL) and DBU (224 µL) was added. The reaction was stirred for 1 h at room temperature and then neutralized with AcOH. The solvent was evaporated in vacuo and the crude residue purified by column chromatography (MeOH/DCM 1:20). The resulting product (21 mg) was identified by LC-MS and subsequently used for the next step. For that, it was dissolved in MeOH (0.25 mL) and 8 (21 mg, 1.00 eq) was added. AcOH (1.00 µL) was added and the reaction was stirred for 2 h at room temperature. Next the reaction was cooled to 0 °C and NaBH4 (2.87 mg, 75.9 µmol, 2.60 eq) was added. The reaction was stirred at 0 °C for 15 min before the solvent was evaporated. Product formation was confirmed by LC-MS. The residue was dissolved in DCM (0.9 mL) and TFA (0.1 mL) was added at 0 °C. The reaction was stirred at 0 °C for 20 min before it was neutralized with NEt3. After evaporation of the solvent, the product was detected by LC-MS and then re-dissolved in EtOAc (0.5 mL). HF • pyridine (15 µL) was added and the solution was stirred for 1 d before the reaction was stopped by addition of TMSOMe (46 µL). The solvent was evaporated, the crude product was identified by LC-MS and dissolved in MeOH (1 mL). NaOMe (27.0 mg, 0.5 mmol) was added and the reaction was stirred for 2 d until LC-MS indicated completion. The mixture was neutralized with NEt3 and the solvent was evaporated in vacuo. The residue was dissolved in water, filtered and purified by reversed phase HPLC (0-10 % buffer B, 45 min) to afford the product 10 (1.30 mg, 2.33 µmol, 5% over 6 steps) as colorless solid.