Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jan 31.
Published in final edited form as: J Am Chem Soc. 2015 Nov 16;137(46):14602–14605. doi: 10.1021/jacs.5b10420

Chemical Mutagenesis of an Emissive RNA Alphabet

Alexander R Rovira 1, Andrea Fin 1, Yitzhak Tor 1,*
PMCID: PMC5281058  NIHMSID: NIHMS842522  PMID: 26523462

Abstract

An evolved fluorescent ribonucleoside alphabet comprising isomorphic purine (tzA, tzG) and pyrimidine (tzU, tzC) analogues, all derived from isothiazolo[4,3-d]pyrimidine as a common heterocyclic core, is described. Structural and biochemical analyses illustrate that the nucleosides, particularly the C-nucleosidic purine analogues, are faithful isomorphic and isofunctional surrogates of their natural counterparts and show improved features when compared to an RNA alphabet derived from thieno[3,4-d]-pyrimidine. The restoration of the nitrogen in a position equivalent to the purines’ N7 leads to “isofunctional” behavior, as illustrated by the ability of adenosine deaminase to deaminate tzA as effectively as adenosine, the native substrate.


Fluorescent nucleoside analogues, which have been developed to address the nonemissive nature of the nucleobases found in DNA and RNA, have found great utility in biophysical analysis and discovery assays.1 In addition to favorable photophysical features, a key element that dictates their utility and ultimate use is their structural similarity to the native counterparts. Characterized as their isomorphic nature, researchers have tried to maximize this trait by minimizing changes to the Watson–Crick (WC) pairing face while introducing fluorescence-enhancing electronic perturbations.2 The impact of these alterations on the physical properties (e.g., tautomerization) and photophysics (excited state level and dynamics) are, however, frequently unpredictable and ultimately empirically assessed.

We have previously completed the first emissive RNA alphabet,3 comprising pyrimidine and purine analogues, all fundamentally derived from thieno[3,4-d]-pyrimidine as a common heterocyclic nucleus (Figure 1). While highly emissive and valuable as illustrated by several applications,4 a limitation of the previously reported thieno analogues is the lack of the basic nitrogen, corresponding to N7 in the purine skeleton (Figure 1).5 As many biomolecular interactions of purine nucleosides and nucleotides rely on the basicity and coordinating ability of this position, we have sought to reinstall this functionality into a new isomorphic RNA alphabet, with higher structural and electronic similarity to the native purines. Herein we report the synthesis, photophysical analysis, and performance of a next generation alphabet based on an isothiazolo[4,3-d]pyrimidine core. This atomic mutagenesis reinstates the basic moiety at the native position yielding analogues that better mimic a purine (Figure 1). This chemically evolved isomorphic alphabet was found to not only have unique photophysical features but also to be highly “isofunctional”, illustrated by the ability of adenosine deaminase (ADA) to deaminate the adenosine analogue as effectively as adenosine, the native substrate.

Figure 1.

Figure 1

Evolution of an emissive RNA alphabet.

While standard conditions provided the isothiazole uridine and cytidine analogues tzU and tzC, respectively,6 common approaches to forming the carbon–carbon glycosidic bonds of the adenosine and guanosine analogues failed under a wide variety of conditions.7,8 A less common approach was required to form the desired connectivity.9 A strategy starting from a ribofuranose-derived precursor as opposed to late-stage glycosylation with the nucleobase and desired sugar was conceived. To construct the C-glycosylated isothiazole ring, a ribofuranose derivative substituted with a primary thiol 3 was synthesized (Scheme 1), starting from a known benzyl-protected precursor, which was subjected to Swern oxidation conditions to give 1.10 Treatment of 1 with diiodomethane and methyllithium furnished the primary halide 2,11 which was reacted with potassium thioacetate and reduced to give the primary thiol 3 (Scheme 1).

Scheme 1. Synthesis of the Sugar Precursora.

Scheme 1

aReagents and conditions: (a) CH2I2, MeLi, toluene, −78 °C, 1 h, 67%. (b) Potassium thioacetate, DMF, rt, 6 h, 74%. (c) Et2O, LiAlH4, 0 °C to rt, 1 h, >90%.

Reactions of 3 with both ester and amide-substituted N-tosyl derivatives 4a and 4b furnished the cyclized 5a and 5b, respectively, in good yields (Scheme 2).12 To set the stereocenter at the anomeric carbon, reduction of 5a and 5b with triethylsilane and BF3·OEt2 was found to be the most effective, yielding key precursors 6a and 6b, respectively (Scheme 2).13 Only one diastereomer was isolated and found to be the desired one (see below). With this key substrate in hand, the synthesis of the protected guanosine analogue 7 was accomplished using a mild two-step, one-pot reaction with an isothiocyanate precursor.14 After cleavage of the carbamate, deprotection using 1,2-ethanedithiol and BF3·OEt2 yielded the final nucleoside tzG.15,16 Synthesis of the adenosine analogue was accomplished via initial construction of the inosine analogue 8 using triethyl orthoformate. This was subsequently converted to the thioamide 9 by treatment with P2S5 in pyridine followed by methanolic ammonia to give the protected final product, which was then subjected to the same deprotection conditions to afford tzA (Scheme 2).

Scheme 2. Synthesis of Purine Analoguesa.

Scheme 2

aReagents and conditions: (a) 3, EtOH, morpholine, 5 h, 0 °C to rt, 67% (5a), 79% (5b). (b) Et3SiH, BF3·OEt2, DCM, −78 °C to rt, 4 h, 63% (6a), 79% (6b). (c) i. EtOC(O)NCS, CH3CN, rt, 4 h; ii. EDCl, HMDS rt, 48 h, 63% (d) i. 1 M NaOH, MeOH, 65 °C; ii. HSCH2CH2SH, BF3·OEt2, DCM, rt, 48 h, 59%. (f) CH(OEt)3, Ac2O, 120 °C, 16 h, 69%. (g) P2S5, pyridine, 65 °C, 2 h. (h) i. NH3, MeOH, 70 °C, 16 h; ii. HSCH2CH2SH, BF3·OEt2, DCM, rt, 72 h, 32% (for steps g and h).

Crystal structure determination confirmed the proposed structure and anomeric configuration of the modified ribonucleosides (Figure 2 and Tables S1–S4). In the solid state, the purine analogues tzA and tzG showed an anti-orientation at the glycosidic linkages, while the pyrimidines tzU and tzC were found to be in the syn-orientation (Figure 2). Rewardingly, analysis of tzG’s crystal packing pattern shows pairing through both the WC and Hoogsteen faces, which is identical to the pattern seen for guanosine in the solid state (Figure 3).17 This intermolecular hydrogen bonding arrangement, illustrating the restoration of a “functional” Hoogsteen face and “N7”, suggests that tzG is likely to share G’s H-bonding and tautomeric preferences.

Figure 2.

Figure 2

X-ray crystal structures of isothiazolo[4,3-d]pyrimidine analogues: (a) tzA, (b) tzG, (c) tzU, and (d) tzC.

Figure 3.

Figure 3

Comparison of intermolecular H-bonding seen in the crystal structures of (a) tzG and (b) G.

The fundamental spectroscopic properties, the sensitivity toward environmental polarity, and the spectroscopically derived pKa values of the modified nucleoside analogues are listed in Table 1. The ground-state absorption spectra in aqueous solution displayed bathochromic-shifted maxima compared to the corresponding native nucleosides ranging from 312 to 338 nm for tzU to tzA, respectively (Figure 4a). Visible emission maxima ranging from 392 nm (for tzU) to 459 nm (for tzG) were observed for all modified nucleosides upon excitation at their maxima. The emission quantum yield of the purine analogues, 0.25 for tzG and 0.05 for tzA, was higher than the pyrimidine analogues, with 0.01 for tzU and 0.05 for tzC in water.

Table 1.

Photophysical Properties of Isothiazolo[4,3-d]pyrimidine Nucleoside Analogues

pKac

solvent λabs (ε)a λem (Φ)a Φε Stokes shifta polarity sensitivityb abs em
tzA water 338 (7.79) 410 (0.05) 413 5.23 27.7 4.25 3.29
dioxane 342 (7.42) 409 (0.03) 193 4.76
tzC water 325 (5.45) 411 (0.05) 289 6.42 10.5 2.94 2.46, 10.38
dioxane 333 (5.03) 419 (0.04) 181 6.14
tzG water 333 (4.87) 459 (0.25) 1203 8.27 102.0 3.55, 8.51 9.88
dioxane 339 (4.65) 425 (0.17) 539 6.01
tzU water 312 (5.17) 392 (0.01) 41 6.53 45.4 2.25, 8.88 8.94
dioxane 314 (5.20) 377 (<0.01) 21 5.36
tzI water 316 (7.63) 377 (0.01) 46 5.13 12.7 9.26 7.83
dioxane 315 (6.62) 372 (<0.01) 26 4.79
a

λabs, ε, λem, and Stokes shift are reported in nm, 103 M−1 cm−1, nm, and 103 cm−1 respectively. All photophysical values reflect the average of at least three independent measurements.

b

Sensitivity to solvent polarity reported in cm−1/(kcal mol−1) is equal to the slope of the linear fit in Figure 4b.

c

pKa values reflect the average over three independent measurements and are equal to the inflection point determined by the fitting curves in Figure 4c,d. See Supporting Table S5 for expanded data including experimental errors.

Figure 4.

Figure 4

(a) Absorption (dashed lines) and emission (solid lines) spectra of tzA (red), tzC (green), tzG (orange), and tzU (purple) in water. (b) Stokes shift correlation versus solvent polarity (ET(30)) of water/dioxane mixtures for tzA (red), tzC (green), tzG (orange), and tzU (purple). (c) Absorption maxima and (d) emission maxima variation versus pH for tzA (red), tzC (green), tzG (orange), and tzU (purple).

The absorption spectra taken in dioxane showed batho and hypochromic shifts for tzA, tzC, and tzG in comparison to the aqueous solutions, while no significant variations were observed for tzU. The emission intensity is sensibly lower in dioxane. The fluorescence maxima displayed a remarkable hypsochromic shift for tzG and tzU, a slight blue-shift for tzA, and a bathochromic shift for tzC. This suggests a charge-transfer character of their excited states (Figure S5). This is manifested for all nucleosides, albeit to different extents, in their responsiveness toward polarity changes, as seen by the linear correlations between the measured Stokes shifts and microscopic solvent polarity parameters (Figure 4b, Table 1).

All new nucleosides display sensitivity toward pH variations, thus facilitating the extraction of pKa values (Figure 4c,d and Table 1). The deprotonation of tzU between pH 8 and 11 is characterized by a red shift of its absorption maximum yielding a pKa value of 8.9, which is comparable to the values reported for N3 deprotonation in uridine (pKa 9.20–9.25).18 tzC is the least responsive to pH changes, displaying minor blue shifts in both the absorption and emission spectra (Figure 4c,d), and yielding two pKa values (Table 1). tzA was characterized by a bathochromic shift both in its absorption and emission maxima upon deprotonation of N1 (pKa = 4.25 and 3.29, respectively) in close proximity to the reported values for adenosine (pKa 3.6– 4.2).19 pH titration of tzG showed two distinct red-shifted transitions of the absorption maximum and two different isosbestic points at 325 and 333 nm (Figure S2), assigned to deprotonation of N7 and N1 (pKa = 3.55 and 8.51, respectively). These values correlate well with the reported values for guanosine (pKa = 3.2–3.3 and 9.2–9.6, respectively).20 Taken together, these observations not only illustrate the responsiveness of these nucleosides, but further indicate the ability of the isothiazole ring to mimic the basic imidazole moiety in the native purines.

To demonstrate the utility of the new analogues and the impact of restoring the basic N7 in evolving the thieno alphabet into the isothiazolo one, we selected to probe the enzymatic deamination of adenosine to inosine. This important metabolic transformation is catalyzed by adenosine deaminase (ADA).21 We have previously reported the ability of ADA to recognize and deaminate thA, the thieno[3,4-d]-pyrimidine-based analogue of the naturally occurring adenosine, to thI (Figure 5a), the corresponding inosine derivative. While of significance in and of itself, the enzymatic conversion of thA to thI was, however, approximately 20-times slower compared to that of adenosine.4b Since ADA has been crystallographically shown to form a H-bond to the N7 of its substrate,21a,b we hypothesized that restoration of this functionality in tzA, the new adenosine surrogate, should facilitate its deamination compared to thA.

Figure 5.

Figure 5

(a) Deamination of thA and tzA. (b) Enzymatic deamination of A to I (black), tzA to tzI (red), and thA to thI (gray) with ADA monitored by real-time absorption at 260, 340, and 340 nm, respectively. Inset: zoomed in region between 0 and 300 s. (c) Enzymatic deamination of tzA to tzI by ADA monitored by real-time absorption (red) at 340 nm and real time emission (light purple) at 410 nm (excitation at 322 nm). Inset: HPLC relative peak area variation at different time-points for tzA (red) and tzI (blue) monitored at 340 nm.

Steady state absorption and emission spectra, taken at defined time-intervals upon addition of ADA to tzA, showed a fast and efficient conversion of tzA to tzI (Figure S6). Real-time continuous measurements, relying on the photophysical differences between tzA and tzI, thA and thI, and A and I,4b,22 show that, while the deamination reaction of thA is indeed sluggish, ADA deaminates tzA to tzI at the same rate as it deaminates A to I (Figure 5b). The reaction half-life for tzA deamination, calculated assuming a pseudo-first order reaction, was comparable to the one observed for adenosine, the native substrate (t1/2 = 39 and 57 s, respectively) and substantially shorter than the one obtained for thA (t1/2 = 818 s). This remarkable initial deamination rate of tzA by ADA, substantiating our hypothesis, was confirmed by HPLC analyses (Figure 5c inset, Figure S8). Taken together, these observations highlight the improved functionality of the isothiazolopyrimidine over the thienopyrimidine core.

In summary, we introduce a second-generation family of emissive nucleoside analogues, based on an isothiazolopyrimidine scaffold. This atomic mutation results in higher isomorphicity and significantly improved functionality when compared to the thienopyrimidine-based RNA alphabet. The presence of the isothiazole core with its nitrogen in the equivalent position to N7 of the native purines, restores the Hoogsteen face, as well as the basicity and native H bonding ability of these surrogates, as illustrated structurally (for tzG) and biochemically (for tzA deamination by ADA). These observations, indicating that the newly introduced purine surrogates display improved structural and functional characteristics, are of significance since very few emissive, isomorphic, and non-perturbing purine analogues have so far been made and biophysically exploited.

Supplementary Material

SI

Acknowledgments

We thank the National Institutes of Health for generous support (GM 069773), the Chemistry and Biochemistry MS Facility, and the UCSD X-ray crystallography Facility.

Footnotes

ASSOCIATED CONTENT

Supporting Information

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.5b10420.

Synthetic details, photophysical data, enzymatic protocols and HPLC traces (PDF)
  • X-ray crystallographic data for tzG (CIF)
  • X-ray crystallographic data for tzC (CIF)
  • X-ray crystallographic data for tzU (CIF)
  • X-ray crystallographic data for tzA (CIF)

The authors declare no competing financial interest.

REFERENCES

  • 1.(a) Sinkeldam RW, Greco NJ, Tor Y. Chem. Rev. 2010;110:2579–2619. doi: 10.1021/cr900301e. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Wilhelmsson MQ. Rev. Biophys. 2010;43:159–183. doi: 10.1017/S0033583510000090. [DOI] [PubMed] [Google Scholar]; (c) Hawkins ME. Cell Biochem. Biophys. 2001;34:257–281. doi: 10.1385/CBB:34:2:257. [DOI] [PubMed] [Google Scholar]; (d) Wilson JN, Kool ET. Org. Biomol. Chem. 2006;4:4265–4274. doi: 10.1039/b612284c. [DOI] [PubMed] [Google Scholar]; (e) Okamoto A, Saito Y, Saito I. J. Photochem. Photobiol., C. 2005;6:108–122. [Google Scholar]; (f) Dodd DW, Hudson RHE. Mini-Rev. Org. Chem. 2009;6:378–391. [Google Scholar]; (g) Kimoto M, Cox RS, III, Hirao I. Expert Rev. Mol. Diagn. 2011;11:321–331. doi: 10.1586/erm.11.5. [DOI] [PubMed] [Google Scholar]; (h) Rist MJ, Marino JP. Curr. Org. Chem. 2002;6:775–793. [Google Scholar]; (i) Wierzchowski J, Antosiewicz JM, Shugar D. Mol. BioSyst. 2014;10:2756–2774. doi: 10.1039/c4mb00233d. [DOI] [PubMed] [Google Scholar]
  • 2.(a) Gaied NB, Glasser N, Ramalanjaona N, Beltz H, Wolff P, Marquet R, Burger A, Mely Y. Nucleic Acids Res. 2005;33:1031–1039. doi: 10.1093/nar/gki253. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Nadler A, Strohmeier J, Diederichsen U. Angew. Chem., Int. Ed. 2011;50:5392–5396. doi: 10.1002/anie.201100078. [DOI] [PubMed] [Google Scholar]; (c) Dierckx A, Miannay F-A, Gaied NB, Preus S, Björck M, Brown T, Wilhelmsson LM. Chem. - Eur. J. 2012;18:5987–5997. doi: 10.1002/chem.201103419. [DOI] [PubMed] [Google Scholar]; (d) Dumat B, Bood M, Wranne MS, Lawson CP, Foller Larsen A, Preus S, Streling J, Gradén H, Wellner E, Grøtli M, Wilhelmsson LM. Chem. - Eur. J. 2015;21:4039–4048. doi: 10.1002/chem.201405759. [DOI] [PubMed] [Google Scholar]; (e) Greco NJ, Tor Y. J. Am. Chem. Soc. 2005;127:10784–10785. doi: 10.1021/ja052000a. [DOI] [PubMed] [Google Scholar]
  • 3.Shin D, Sinkeldam RW, Tor Y. J. Am. Chem. Soc. 2011;133:14912–14915. doi: 10.1021/ja206095a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.(a) Liu W, Shin D, Tor Y, Cooperman BS. ACS Chem. Biol. 2013;8:2017–2023. doi: 10.1021/cb400256h. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Sinkeldam RW, McCoy LS, Shin D, Tor Y. Angew. Chem., Int. Ed. 2013;52:14026–14030. doi: 10.1002/anie.201307064. [DOI] [PMC free article] [PubMed] [Google Scholar]; (c) McCoy LS, Shin D, Tor Y. J. Am. Chem. Soc. 2014;136:15176–15184. doi: 10.1021/ja5039227. [DOI] [PMC free article] [PubMed] [Google Scholar]; (d) Otomo H, Park S, Yamamoto S, Sugiyama H. RSC Adv. 2014;4:31341–31344. [Google Scholar]; (e) Park S, Otomo H, Zheng L, Sugiyama H. Chem. Commun. 2014;50:1573–1575. doi: 10.1039/c3cc48297a. [DOI] [PubMed] [Google Scholar]; (f) Sholokh M, Sharma R, Shin D, Das R, Zaporozhets OA, Tor Y, Mely Y. J. Am. Chem. Soc. 2015;137:3185–3188. doi: 10.1021/ja513107r. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Mizrahi RA, Shin D, Sinkeldam RW, Phelps KJ, Fin A, Tantillo DJ, Tor Y, Beal PA. Angew. Chem., Int. Ed. 2015;54:8713–8716. doi: 10.1002/anie.201502070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.The synthesis of the pyrimidine analogues is described in the Supporting Information (Scheme S5).
  • 7.(a) Coté GL, Flitsch S, Ito Y, Kondo H, Nishimura S, Yu B, Fraser-Reid BO, Tatsuta K, Thiem J. Glycoscience: Chemistry and Chemical Biology. 2. Berlin Heidelberg: Springer-Verlag; 2008. [Google Scholar]; (b) Iddon B. Heterocycles. 1995;41:533–593. [Google Scholar]; (c) Stambasky J, Hocek M, Kocovsky P. Chem. Rev. 2009;109:6729–6764. doi: 10.1021/cr9002165. [DOI] [PubMed] [Google Scholar]
  • 8.Attempts to glycosylate the nucleobase using palladium coupling chemistry, grignard formation, alkyllithium chemistry, and standard Vorgbruggen glycosylation conditions were all unsucessful.
  • 9.Wamhoff H, Berressem R, Nieger M. J. Org. Chem. 1993;58:5181–5185. [Google Scholar]
  • 10.See Supporting Information (Scheme S1).
  • 11.Bessieres B, Morin C. Synlett. 2000:1691–1693. [Google Scholar]
  • 12.Gewald K, Bellmann P. Liebigs Ann. Chem. 1979;1979:1534–1546. [Google Scholar]
  • 13.van Rijssel ER, van Delft P, Lodder G, Overkleeft HS, van der Marel GA, Filippov DV, Codée JDC. Angew. Chem., Int. Ed. 2014;53:10381–10385. doi: 10.1002/anie.201405477. [DOI] [PubMed] [Google Scholar]
  • 14.Lecoutey C, Fossey C, Rault S, Fabis F. Eur. J. Org. Chem. 2011;2011:2785–2788. [Google Scholar]
  • 15.Fuji K, Ichikawa K, Node M, Fujita E. J. Org. Chem. 1979;44:1661–1664. [Google Scholar]
  • 16.Seley KL, Zhang L, Hagos A, Quirk S. J. Org. Chem. 2002;67:3365–3373. doi: 10.1021/jo0255476. [DOI] [PubMed] [Google Scholar]
  • 17.(a) Additionally, an overlay of the purine analogues and natural nucleosides may be found on Figure S1. Marsh RE, Bugg CE, Thewalt U. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1970;B26:1089–1101. doi: 10.1107/s0567740870003667.
  • 18.(a) Simpson RB. J. Am. Chem. Soc. 1964;86:2059–2065. [Google Scholar]; (b) Luyten I, Pankiewicz KW, Watanabe KA, Chattopadhyaya J. J. Org. Chem. 1998;63:1033–1040. [Google Scholar]
  • 19.(a) Christensen JJ, Rytting JH, Izatt RM. Biochemistry. 1970;9:4907–4913. doi: 10.1021/bi00827a012. [DOI] [PubMed] [Google Scholar]; (b) Kapinos LE, Operschall BP, Larsen E, Sigel H. Chem. - Eur. J. 2011;17:8156–8164. doi: 10.1002/chem.201003544. [DOI] [PubMed] [Google Scholar]
  • 20.(a) Bundari S. The Merck Index. 12th. Whitehouse Station, NJ: Merck and Co., Inc; 1996. [Google Scholar]; (b) Sigel H, Massoud SS, Corfù NA. J. Am. Chem. Soc. 1994;116:2958–2971. [Google Scholar]; (c) Thapa B, Schlegel HB. J. Phys. Chem. A. 2015;119:5134–5144. doi: 10.1021/jp5088866. [DOI] [PubMed] [Google Scholar]; (d) Kampf G, Kapinos LE, Griesser R, Lippert B, Sigel H. J. Chem. Soc., Perkin Trans. 2002;2:1320–1327. [Google Scholar]
  • 21.(a) Kinoshita T, Nishio N, Nakanishi I, Sato A, Fujii T. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2003;59:299–303. doi: 10.1107/s090744490202190x. [DOI] [PubMed] [Google Scholar]; (b) Wilson DK, Rudolph FB, Quiocho FA. Science. 1991;252:1278–1284. doi: 10.1126/science.1925539. [DOI] [PubMed] [Google Scholar]; (c) Kinoshita T, Nakanishi I, Terasaka T, Kuno M, Seki N, Warizaya M, Matsumura H, Inoue T, Takano K, Adachi H, Mori Y, Fujii T. Biochemistry. 2005;44:10562–10569. doi: 10.1021/bi050529e. [DOI] [PubMed] [Google Scholar]
  • 22.See Supporting Information (Figure S6).

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SI

RESOURCES