Second-Generation Chiral Amino Acid Derivatives Afford High Stereoselectivity and Stability in Aqueous RNA Acylation

Abstract

Activated acyl species have proven versatile in the esterification of 2′–OH groups in RNA, enabling structure mapping, caging, profiling, and labeling of the biopolymer. Nearly all reagents developed for this reaction have been achiral; however, a recent study reported that simple chiral amino acid acylimidazole derivatives could yield diastereoselective reactions at RNA 2′–OH in water, enabling up to 4:1 selectivity in screening. Here, we investigated the effect of steric bulk on the stereoselectivity of RNA reaction and on the stability of adducts with a library of 36 chiral acylimidazole scaffolds with increasing steric demand. The results document the highest stereoselectivity yet achieved in RNA acylation reactions, with as high as >99:1 diastereoselectivity at >70% conversion. Also notably, the bulky adducts were found to have markedly improved stability on RNA.

Graphical Abstract

INTRODUCTION

The RNA 2′–OH group contributes critically to the unique chemical, structural, and biological properties of the biopolymer. It helps to determine the ribose sugar conformation, which in turn influences the overall right-handed A-form shape of the helix.¹ The group is also essential to RNA splicing mechanisms,² and it defines the site of instability in RNA, as thermal and enzymatic chain cleavage mechanisms begin with this nucleophilic hydroxyl group.³ Because the 2′–OH group occurs at virtually every position of RNA, it also can act as a useful handle for analysis and modification of the polymer.^4,5 The unusually low pK_a of this group enables it to act as a nucleophile toward activated carbonyl and sulfonyl species at neutral pH.^6–13 Appropriately designed acyl and alkyl acylimidazoles and aryl isatoic anhydrides are widely used to acylate this group in trace yields for mapping the folding and interactions of RNA,^4,5 and reagents that can provide high-yield reactions at 2′–OH offer useful methods for conjugation of the biopolymer.^11,14

The natural d-ribose monomeric scaffold of RNA presents a chiral substrate for reactions at 2′–OH, and complex folding of the biopolymer can create local macrochiral environments that may further magnify the effects of asymmetry in small-molecule interactions. Such asymmetry was of interest in studies of prebiotic chemistry several decades ago, investigating origins of amino acid charging of tRNAs (tRNAs).^15,16 Reactions of activated amino acids with dinucleotides were reported, resulting in up to 2:1 diastereoselectivity and up to 7% yields.^17,18 Interestingly, the preferred stereochemistry for amino acids reacting with RNA was found to be the d-enantiomer, inconsistent with the existing biological l-amino acids, which are enzymatically conjugated to tRNA termini by tRNA synthetases.

If highly stereoselective reagents and practical high-yield reactions with RNA could be identified, then such reactions can potentially have multiple applications. For example, chiral compounds offer the possibility of elevated signal over background in RNA structure mapping.¹⁹ In addition, they may assist in the selection and conjugation of enantiomeric l-RNA aptamers.^20,21 Further, stereoselectivity may provide increased local site selectivity in conjugation reactions.¹⁹ Finally, chiral 2′–OH adducts that are designed to be reversible in biological media²² may well respond differentially to chiral enzymes and chiral small molecules in the cell. Until recently, there were no reports of stereoselectivity at preparatively useful yields in small-molecule covalent reactions with RNA. However, in a recent study, we reported the initial finding that RNA can react differentially with enantiomers of simple chiral acylimidazole compounds.¹⁹ Amino acid and alkoxy acid imidazolide derivatives were tested in aqueous buffers, and up to 3.9:1 diastereoselectivity was seen in preparative-yield reactions of an alanine derivative with unfolded single-stranded RNA. Yet higher selectivity (up to 16:1) was observed for local positions in folded RNA targets, where the chiral environment can be more rigidly defined.¹⁹ While the early results in the stereoselective acylation of RNA were promising, it remained unclear whether additional structural design of activated acyl species could lead to higher stereoselectivity or whether increasing steric bulk around the chiral center, as a strategy for achieving this, might adversely affect reactivity.

In addition to acylation of RNA, the reversibility of this acylation has also been explored recently.²² The premature loss of acyl adducts by hydrolysis can result in unwanted loss of protection of RNA, and it can shorten the useful life span of conjugates. Conversely, active removal of substituents by nucleophiles can be used to activate biochemical and biological function that are blocked by this substitution,^23–25 for example, by removal of nuclease-protective acyl groups on RNA to enable efficient translation. Given the early nature of the work with chiral RNA adducts, it remained unknown how increasing the steric bulk of stereoselective reagents might affect hydrolysis rates or nucleophilic reversal of adducts. Any differences among adducts might also enable the orthogonal removal of groups.

RESULTS AND DISCUSSION

Reagent Design.

We investigated these issues by preparing and studying a set of new chiral acylating agents with an increasing steric demand near the activated acyl group. The 36 compounds studied here ((R/S) 1–18) are shown in Figure 1. A previous study identified N,N-dimethylalanine analogue (R)-8 as a selective chiral species for naturally substituted RNAs.¹⁹ In early screening, this compound exhibited 3.9:1 (R/S) diastereoselectivity in reactions with single-stranded RNA. We hypothesized that a potentially effective mechanism for enhancing the stereoselectivity of analogues of this compound could be to increase steric size and/or alter the geometry of one of the groups near the reactive acyl group.²⁶ Note that we employed dialkyl or acyl substitution on the amine group in order to avoid polymerization, which we observed when testing activated unprotected amino acid compounds reported several decades ago¹⁸ (see Supporting Information (SI), Figure S1).

Figure 1.

Diastereoselective reactions of chiral acylating reagents with RNA. (A) The 2′–OH group of RNA is located within a chiral sugar and helix, but although it is a hindered secondary alcohol, it has elevated reactivity in water. (B) Prior acylating reagents showed moderate selectivity and low adduct hydrolytic stability in initial screens. (C) New second-generation reagents with greater steric bulk show unprecedented diastereoselectivity and elevated ester adduct stability. (D) Structures of 36 chiral acylimidazole compounds tested.

Following this strategy, we prepared a set of new compounds (Figure 1) based on amino acids or α-alkoxy acids with increased alkyl or aryl bulk, either at the α-carbon or at the N,O heteroatoms. For direct comparison with the earlier report,¹⁹ we included the reported enantiomer pair (R,S)-8, the least sterically encumbered structure in the new set. For initial screening, the compounds were converted to activated acylimidazole derivatives as stock solutions in anhydrous dimethyl sulfoxide (DMSO) by a 1 h reaction with carbonyldiimidazole (CDI) (23 °C). Note that this also produces 1 equiv of imidazole as a byproduct, which was also present in reactions. The compounds were initially screened for reactivity and stereoselectivity with a 20 nucleotide (nt) single-stranded test RNA at 10 mM in pH 7.5 MOPS buffer (see the SI). Reactions were performed in most cases at 0 °C for 2 h. We initially tested 2–4 concentrations of each compound (0.1–100 mM) to determine an appropriate concentration that would achieve moderate (ca. 20–80%) conversions; each enantiomer pair was screened at identical concentrations so that relative stereoselectivity could be assessed. Reaction conversions (unmodified RNA conversion to RNA with at least one modification) were evaluated by quantitative matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry, which yields linear responses with small modified RNAs.¹³

Reactivity Screening.

Results of the screening (Table 1) revealed wide ranges of reactivity and selectivity among the 36 compounds tested. Stereoselectivities varied from near zero (~1:1 R/S ratio of conversion; e.g. compounds 3,4) to relatively high (>4:1; compounds 12,14,16–18). These latter stereoselective amino acid derivatives all reacted preferentially as R enantiomers, similar to the prior alanine derivative 8.¹⁹ The highest stereoselectivity was observed for compound 17, at 25.2:1 (R/S), more than 6-fold greater than that observed for lead compound 8. Reactivities also varied greatly, with the most reactive compounds (3,6–8,10,13,14) achieving the target range of conversion at remarkably low concentrations of 100 μM. This is especially noteworthy, given that many RNA acylation experiments employ reagents at ca. 10–200 mM even to produce trace yields for structure mapping.^4,27 Other compounds, typically sterically bulky amino acid-derived or uncharged alkoxy derivatives, required higher reagent concentrations (100 mM; e.g., 1,2,5). Strikingly, N-acetylated alanine compound 4 was more than 200-fold less reactive than N-dimethylalanine compound 8, suggesting the importance of a protonated amine group for achieving high reactivity. Interestingly, 4 also displayed an almost complete loss of stereoselectivity compared with 8. Finally, all compounds were also tested with a DNA substrate having the same sequence as the test RNA, revealing little or no reaction with the DNA (Figure S3). As this DNA contains the same amine and terminal hydroxyl groups as the RNA but lacks 2′–OH groups, this strongly suggests 2′–OH groups in the RNA as the sites of reaction rather than nucleophilic sites on the bases or the 5′/3′ hydroxyl termini. It is also important to note that these early measurements of diastereoselectivity are averages over the whole RNA, which may mask local stereoselectivities that differ. However, in a previous study, single-stranded RNA did not show marked differences in stereoselectivity along the strand for compound 8.¹⁹

Table 1.

Reaction Conversion and Diastereoselectivity for Chiral Scaffolds 1–18 during Screening for Acylation of Single-Stranded RNA^a

structure	compound	concentration (mM)	conversion (%)b	diastereoselectivity (R/S)
	(R) 1	100	63	1.54
	(S) 1		41
	(R) 2	100	36	0.56
	(S) 2		64
	(R) 3	0.10	40	0.92
	(S) 3		47
	(R) 4	20.0	43	1.08
	(S) 4		40
	(R) 5	100	46	1.77
	(S) 5		26
	(R) 6	0.10	43	0.96
	(S) 6		45
	(R) 7	0.10	65	0.94
	(S) 7		69
	(R) 8c	0.10	98	3.92
	(S) 8c		25
	(R) 9	0.50	44	2.75
	(S) 9		16
	(R) 10	0.10	77	1.97
	(S) 10		39
	(R) 11	1.00	43	3.07
	(S) 11		14
	(R) 12	5.00	92	5.41
	(S) 12		17
	(R) 13	0.10	86	1.83
	(S) 13		47
	(R) 14	0.10	85	4.47
	(S) 14		19
	(R) 15	1.00	85	3.54
	(S) 15		24
	(R) 16	10.0	58	4.46
	(S) 16		13
	(R) 17	5.00	68	25.2
	(S) 17		2.7
	(R) 18	50.0	58	8.78
	(S) 18		6.6

^aConditions: RNA acylation yields for the chiral acylating reagents with a single-stranded RNA. 10 μM RNA was treated with reagent, 20% DMSO in water for 2 h at 0 °C. Reagent concentrations (shown) were adjusted to provide ca. 20–80% conversion.

^bConversion of starting RNA to acylated products as measured by MALDI-TOF MS.

^cData from ref 20.

Having identified several new compounds with elevated reactivity and stereoselectivity, we chose eight (9,12–18) to examine the effects of increasing bulk (Figures 2 and S4). Tests at varied concentrations enabled an assessment of relative reactivity, and time course experiments revealed when preparatively useful conversions to RNA ester derivatives were achieved and when stereoselectivity was maximized. The general trend observed for increasing acyclic substituent size (8,12,17,18; Figure 2C) was that increasing steric bulk slowed reactivity but generally increased diastereoselectivity, culminating in scaffold 17, the most selective of the study. Comparisons of cyclic substitution (13–16) revealed similar trends, with the cyclopentyl derivative 15 showing the greatest diastereoselectivity (Figure S5A,B). Cyclopropyl compound 13, the smallest new compound studied, produced remarkable reactivity, giving >90% conversion at 1 h using only 100 μM reagent, but stereoselectivity (4.4:1) was lower than for the isopropyl case. In contrast, the sterically largest compound, 3-pentyl compound 17, required higher concentrations but displayed 84% conversion after 1 h (10 mM) and the highest selectivity of the group, at 16:1 under those conditions. At 5 mM, the reaction of 17 exhibited yet higher 25:1 diastereoselectivity at 64% conversion (Figure 2A,B,E), which is remarkable given the flexible nature of the ssRNA substrate. Finally, adding steric bulk at the amino group (4,8,9–11) rather than the α-carbon generally decreased both the reactivity and selectivity (Figure S5C,D). Alkoxy compounds were considerably less reactive than most alkylamino compounds and provided lower stereoselectivity, but with a general preference for the opposite (S) enantiomer as compared with the amino acids (Table S1).

Figure 2.

Effect of conditions and structure on reactions of selected chiral compounds with single-stranded RNA. (A) Effect of concentration on RNA conversion for the two enantiomers of compound 17. Data are averages from at least three experiments; error bars represent standard deviations. (B) Time course of conversion for the enantiomers of 17 (10 mM reagent). Data are averages from at least three experiments; error bars represent standard deviations. (C, D) Effect of increasing steric bulk on conversion and stereoselectivity in compounds 8, 12, 17, and 18. See Figure S4 for the original data and reproducibility. (E) Representative MALDI-TOF mass spectra for single-stranded RNA products of enantiomers of compound 17. Conditions: 10 mM RNA, 0 °C, 2 h unless listed otherwise.

The above tests measured chiral selectivity in reactions with an unconstrained and flexible single-stranded RNA. To evaluate the selectivity of a second-generation reagent in a constrained setting, we evaluated reactions of the sterically largest and most selective compound 17 with an RNA having a small reactive single-nucleotide bulge loop, which can provide a more chirally constrained environment,^19,28 induced by a DNA helper.¹⁴ Previous work has shown that RNA acylation reactions occur selectively with nucleotides in loops over duplex due to greater steric accessibility in the former, with small loops being the most reactive.²⁸ The results with 10 mM reagents showed that the (R) enantiomer provided preparative-level yields (71%) of a single adduct, which we ascribe to reaction at the bulge loop nucleotide (Figures 3A and S6). A control experiment with a fully complementary DNA (no induced loop) gave little or no conversion, providing further evidence of the localized site of the reaction. Remarkably, no measurable acylation of this RNA was seen for the (S) enantiomer; thus, the diastereoselectivity is at least 99:1 under these conditions.

Figure 3.

Effects of steric bulk and high diastereoselectivity in reactions of reagents (R or S) 17 with folded RNAs. (A) Local preparative-level acylation of a small RNA with an induced 1-nt bulge loop; essentially complete diastereoselectivity for (R) (>99:1) is observed at conversions of 71%. See Figure S6 for primary data and the control experiment. (b) Trace-level reactions for RNA secondary structure mapping of flavin mononucleotide riboswitch with enantiomers of chiral reagent 17, showing expected bands corresponding to loops in FMN RNA for (R)–17 (1 mM). Note the much higher reactivity of the R enantiomer over S (see the SI for experimental details). Established structure-probing nicotinyl acylimidazole reagent (NAI) was tested for comparison. At right is the established folded secondary structure of FMN RNA,²⁹ with unpaired reactive loops color-coded in red (blue indicates primer binding site).

Adduct Stability.

RNA-acylating compounds that are highly reactive often achieve this reactivity via an electron-deficient carbonyl group, typically by substitution with an electron-withdrawing moiety such as in nitroaryl compounds³⁰ or with cationic charge in glycine derivatives.²² This electron deficiency can also lower the stability of the ester adduct on RNA, enhancing reactivity to water. While this can be beneficial with intentionally reversible groups,²² it may be detrimental in applications such as conjugation and structure mapping, where a significant lifetime of the adduct is required during manipulation and analysis. We wondered whether increased steric bulk in some of the reactive compounds studied here might affect the stability of the ester adducts on RNA; we considered the opposing possibilities that steric strain in the adduct might promote aqueous hydrolysis, while on the other hand, steric occlusion near the carbonyl might hinder water attack. To examine these possibilities, we incubated RNA adducts of 15 and 17, with intermediate and high steric bulk respectively, in pH 7.5 phosphate buffer for a week at 23 °C. The results showed (Figures 4A,B and S7) that ester adducts of 15 were ca. 60% hydrolyzed after 1 week, while adducts of bulkier compound 17 were significantly more stable, showing only ca. 15% hydrolysis after 1 week.

Figure 4.

Enhanced stability of ester adducts of sterically bulky compounds on single-stranded RNA. (A, B) Plots of RNA adducts of less-bulky compound 15 (A) and more bulky compound 17 (B), showing greater hydrolysis of esters of 15 over 1 week in phosphate-buffered saline (PBS) (23 °C), while adducts of 17 remain more stable. Graph shows single-stranded RNA adducts incubated over time (see Figure S7 for additional data). (C, D) Tris free base (shown; 100 mM) causes complete loss of relatively unhindered compound 8 after 2 h (23 °C). Blue numerals indicate the number of adducts for selected peaks. (E, F) Effect of Tris on loss of adducts of hindered compound 17 from a single-stranded RNA after 2 h (23 °C), showing reduced loss of adducts due to the steric bulk.

Given a recent observation that nucleophilic amine species can facilitate the removal of certain ester adducts on RNA,²² we were prompted to test whether bulky adducts in the current study might favorably affect the stability to those nucleophiles. We tested multiple amines (e.g., Tris, DABCO, and N-methylimidazole) incubated separately at 100 mM with RNAs containing adducts of bulky compound 17 as well as the less hindered previous compound (R)-8. The data showed (Figures 4C–F and S8) that adducts of nonbulky alanine derivative (R)-8 are highly sensitive to Tris, being completely removed in 2 h (23 °C). However, RNA esters of (R)-17 were found to be resistant to Tris and the other amines under these conditions, showing minor (~11%) or no loss, depending on the amine. Thus, we conclude that the steric bulk of (R)-17 hinders attack of both water and amine nucleophiles, conferring enhanced stability on RNA.

Tests of Elevated Chirality in Structure Mapping.

Our initial study identified a chiral compound that could be used effectively for mapping secondary structure in a folded RNA, taking advantage of the reaction preference for more sterically accessible unpaired nucleotides over those in double-stranded helices.^19,28 Given that several new compounds (particularly 17) show considerably higher diastereoselectivity than the earlier reagent, we explored whether this new, considerably more bulky scaffold could be used for this purpose and whether the chirality might affect reactivity or the ability to distinguish unpaired nucleotides from paired ones. Results with a well-established folded RNA (flavin mononucleotide riboswitch (FMN RNA))²⁹ revealed that at 1 mM the (R) enantiomer of 17 showed clear bands at single-hit levels of reaction necessary for structure probing, while the (S) enantiomer showed very little reactivity (Figure 3B), consistent with the stereoselectivity documented (above) in single-stranded RNA. Importantly, the adducts of (R)-17 caused reverse transcriptase stops, marking sites of reaction, and the bands correspond to known loop sites in the FMN RNA, revealing a reactivity pattern very similar to that of an established achiral reagent (NAI). Interestingly, despite its steric bulk, (R)-17 was able to probe the structure at concentrations considerably lower than those of the achiral species.

CONCLUSIONS

Our studies have succeeded in identifying new, more bulky chiral reagent scaffolds that provide the highest diastereoselectivities yet seen for reaction at RNA 2′–OH groups. This enables elevated selectivity in reactions carried out either stochastically (>100% conversion) in unfolded RNAs or for local site-directed reactions. In single-stranded RNAs, we observed stereoselectivity of 25:1 in reactions carried to preparatively useful conversions using bulky 3-pentylglycine scaffold 17. In a more constrained local acylation reaction, stereoselectivity increased to greater than 99:1. We have further shown that the modification reactions of several of the more bulky reagents can be readily carried out to super-stoichiometric yields, resulting in ester derivatives at over half of the 2′–OH groups in a single-stranded RNA. Such polyacylation has proven useful previously in caging RNA, as well as in protecting RNA from degradation.²² The new results indicate that bulky chiral species may also be useful for such applications and that handedness plays a central role in their reactions on RNA.

Overall, we conclude that chirality should be an important consideration in the design of future reagents for protecting, modifying, and mapping RNA. In addition, the current documentation of highly reactive species such as (R)-13, which reacts in preparative yields with only 100 μM reagent, also suggests new structural designs that may further magnify RNA reactivity in acylimidazole reagents; for example, exploration of new cationic scaffolds is merited by the new findings.

EXPERIMENTAL SECTION

General Analytical Methods.

NMR spectra were recorded at the Stanford University Department of Chemistry NMR facility. Varian 300, 400, and 500 MHz NMR instruments were used to record the ¹H and ¹³C spectra. In general, ¹H NMR solvents were CDCl₃ (7.26 ppm), DMSO-d₆ (2.5 ppm), and 1% MeOH in D₂O (4.79 ppm), and ¹³C NMR solvents were DMSO-d₆ (39.52 ppm) and 1% MeOH in D₂O (49.50 ppm as MeOH). The spectra were analyzed by using MNova software.

SI-MS spectra were recorded with a Waters 2795 HPLC system with dual wavelength UV detector and ZQ single quadrupole MS with electrospray ionization source. MALDI-TOF spectra were recorded with a Bruker Daltonik Microflex MALDI-TOF spectrometer equipped with an N₂ laser. Spectra were recorded in linear negative mode, and samples were plated on an MSP Anchorchip 96 target plate. 0.3 M trihydroxyacetophenone in EtOH (matrix) and 0.1 M aqueous ammonium citrate (comatrix) were mixed in a 2:1 ratio by volume to be used as a matrix mix for MALDI. This mix was always freshly prepared before analysis. After RNA precipitation, the RNA pellet was redissolved in RNase-free water to prepare a 10 μM sample solution. 1 μL of this solution was transferred to the target plate and dried under an Ar stream. 1 μL of the matrix mix was then added directly to the dried sample and completely dried under an Ar stream. The spectral data were then recorded using Flex Control software (Bruker) and analyzed using MNova (Mestrenova).

General Procedure for RNA Reactions with Chiral Acylating Reagents.

In a sterile 200 μL PCR tube, 3.3 μL of buffer concentrate (333 mM MOPS pH 7.5, 333 mM NaCl, 20 mM MgCl₂ in water) was mixed with 4.7 μL of 21.3 μM RNA stock solution. Fresh stocks of chiral acyl reagents were prepared in DMSO; 2 μL of stock solution was added to the RNA reactions. These reactions were incubated for 2 h at 0 °C, and the RNA was subsequently purified by ethanol precipitation. The level of RNA modification was measured by MALDI-TOF M/S and analyzed using MestReNova software.

Ethanol Precipitation of RNA Reactions.

For an RNA reaction with a total volume of 10 μL, 90 μL of RNA precipitation solution (0.33 M NaOAc (pH 5.2) in water containing 0.2 mg/mL glycogen) was added and mixed well. 300 μL of ice-cold absolute ethanol was then added, and the mixture was mixed by vortexing for at least 30 s. After storage at −80 °C overnight, the mixture was centrifuged at 14.8k RPM for 60 min at 4 °C. The supernatant was discarded to obtain a pellet, which was washed with 70% ethanol. The obtained pellet was air-dried for 15 min and subsequently either stored at −80 °C for future use or dissolved in water/PBS for direct use in further experiments.

Induced-Loop RNA Acylation.

100 pmol of RNA and 200 pmol of inducer DNA were heated in folding buffer containing 50 mM NaCl to 95 °C for 3 min and then chilled on ice. To the solution were added 3.3 μL of 3.3× HEPES buffer and 2 μL of 50 mM 17S/R in dry DMSO for the final 10 μL reaction. The mixture was incubated at 0 °C for the desired time (0.5, 1, 2 h). After incubation, 1.25 μL of 0.1 M DTT was added at room temperature (15 min) to quench the reaction. 2 μL of Turbo DNase I (2 U/μL, Invitrogen) and 67 μL of H₂O were directly added to digest the inducer DNA at 37 °C for 1 h. After DNA digestion, the RNA was purified by ethanol precipitation and analyzed by a MALDI-TOF M/S.

In Vitro Transcription.

FMN RNAs were synthesized by in vitro transcription using HiScribe T7 Quick High Yield RNA Synthesis Kit (NEB), following the manufacturer’s protocol. The reaction mixture was incubated at 37 °C overnight. Transcribed RNAs were purified by Quick-RNA MidiPrep kit following the manufacturer’s protocol.

In Vitro RNA Structure Mapping.

To map transcribed FMN RNA structure, 500 ng of FMN RNA was heated in the buffer containing 50 mM NaCl at 75 °C for 3 min and stepped cooled to 25 °C by 0.5 °C/s. 2.5 μL of 4× SHAPE buffer (400 mM MOPs, pH 7.5, 400 mM NaCl, 24 mM MgCl₂) was added to the mixture and incubated at room temperature for 10 min. In a final 10 μL reaction, 2 μL of 5 mM 17S/R in DMSO (+) or DMSO (−) as mock control was added to react at 0 °C for 2 h. The reaction was quenched by 1 μL of 0.1 M DTT and purified by ethanol precipitation.

PAGE Analysis of Reverse Transcriptase (RT) Stops.

4 pmol of FMN RNA was mixed with 4 pmol of Cy5-labeled RT Primer and 0.25 μL of 10 mM dNTP mix (for sequencing lane, ddNTP/dNTP = 8:1), incubated for 5 min at 65 °C, and then immediately chilled on ice for 2 min. 2 μL of 5x First-Strand Buffer, 1 μL of 0.1 M DTT, 0.25 μL of 40 U/μL RiboLock, and 0.25 μL of 200 U/μL SuperScript II were added to the final volume of 10 μL. The reaction was incubated with the following program: 25 °C for 10 min, 42 °C for 50 min, 52 °C for 50 min, and kept at 4 °C. After the reverse transcription reaction, 1 μL of 1 M NaOH was added and incubated at 95 °C for 3 min to remove the RNAs. Then, 11 μL of loading dye (8 M Urea, 0.05% Orange G, 0.05% Bromophenol blue) was added, and the mixture was denatured at 95 °C for 3 min and loaded on a denaturing 8% polyacrylamide gel. Products were separated in a gel in 1× TBE (pH 8.3, Sigma-Aldrich), 25 mA, ~2.5 h. The cDNA gel was visualized by fluorescence imaging.

Data Availability Statement

The data underlying this study are available in the published article and its online supplemetary materials.