The Contribution of the Activation Entropy to the Gas-Phase Stability of Modified Nucleic Acid Duplexes
Tricyclo-DNA (tcDNA) is a sugar-modified analogue of DNA currently tested for the treatment of Duchenne muscular dystrophy in an antisense approach. Tandem mass spectrometry plays a key role in modern medical diagnostics and has become a widespread technique for the structure elucidation and quantification of antisense oligonucleotides. Herein, mechanistic aspects of the fragmentation of tcDNA are discussed, which lay the basis for reliable sequencing and quantification of the antisense oligonucleotide. Excellent selectivity of tcDNA for complementary RNA is demonstrated in direct competition experiments. Moreover, the kinetic stability and fragmentation pattern of matched and mismatched tcDNA heteroduplexes were investigated and compared with non-modified DNA and RNA duplexes. Although the separation of the constituting strands is the entropy-favored fragmentation pathway of all nucleic acid duplexes, it was found to be only a minor pathway of tcDNA duplexes. The modified hybrid duplexes preferentially undergo neutral base loss and backbone cleavage. This difference is due to the low activation entropy for the strand dissociation of modified duplexes that arises from the conformational constraint of the tc-sugar-moiety. The low activation entropy results in a relatively high free activation enthalpy for the dissociation comparable to the free activation enthalpy of the alternative reaction pathway, the release of a nucleobase. The gas-phase behavior of tcDNA duplexes illustrates the impact of the activation entropy on the fragmentation kinetics and suggests that tandem mass spectrometric experiments are not suited to determine the relative stability of different types of nucleic acid duplexes.
KeywordsAntisense Duplex DNA Modified DNA Activation entropy Tandem mass spectrometry
tcDNA in Antisense Therapy
The selective binding of RNA targets encourages the application of tcDNA as antisense oligonucleotide. It is currently being tested for the treatment of Duchenne muscular dystrophy (DMD) . DMD is the most severe form of muscular dystrophies and arises from a lack of the structural protein dystrophin due to random mutations or exon deletion in the dystrophin gene . The goal of antisense therapy is to induce alternative splicing that leads to the expression of an internally deleted but functional protein. Goyenvalle et al. showed that tcDNA promotes significantly higher levels of dystrophin in the mdx mouse model than 2'-OMe-thiophosphate oligonucleotides and morpholino oligomers . In the course of this study, first tandem mass spectrometric analysis of the antisense oligonucleotide revealed unique fragmentation characteristics that are discussed in more detail herein.
MS/MS of Oligonucleotides
Antisense oligonucleotides are designed to exhibit high in vivo stability, which renders them inaccessible for enzyme-based sequencing techniques. Today, tandem mass spectrometry (MS/MS) is the method of choice for the sequencing of structurally modified nucleic acids. Collision-induced dissociation (CID) leads to the scission of the sugar-phosphate backbone at one of the four phosphodiester bonds, giving rise to four ion pairs referred to as a-/w-, b-/x-, c-/y-, and d-/z-ions . The loss of nucleobases constitutes an alternative or concomitant fragmentation pathway of oligonucleotides.
The preferred fragmentation channel is determined by the structure of the nucleic acid. A detailed account of the fragmentation pathways of ESI-generated natural and modified oligonucleotide ions can be found in a recent review . In non-modified DNA, the most frequently observed fragments are a-B- and w-ions. This ion pair is formed in a two-step mechanism triggered by the loss of a nucleobase . In RNA, the primary fragmentation products are c- and y-ions, which has been explained by an alternative fragmentation mechanism involving the 2'-OH group . Moreover, some of the synthetic modifications introduced in antisense oligonucleotides are also known to alter the gas-phase dissociation mechanism. While thiophosphate oligonucleotides fragment in a DNA-like manner , no preferred dissociation channel was observed for 2'-methoxylated oligonucleotides , methylphosphonates , or locked nucleic acids . In fact, the fragmentation patterns point towards the coexistence of multiple mechanisms. Recent tandem mass spectrometric studies on LNA  and homo-DNA  accentuate the impact of the sugar-moiety on the gas-phase dissociation of oligonucleotides. Consequently, understanding the fragmentation of tricyclo-DNA is a key factor for the reliable MS-based sequencing and quantification of the antisense oligonucleotides.
MS/MS of DNA Duplexes
There are two measures of the gas-phase stability of a duplex: the relative signal intensities of duplex and single strands and the activation energy required for fragmentation. The relative signal intensities of matched and mismatched duplexes correlate well with the in-solution stability of the duplex . Alternatively, the stability of a duplex can be evaluated with the E50 value, which corresponds to the activation energy required to dissociate 50% of the duplex. In general, the E50 value and the melting temperature Tm measured in solution correlate for duplexes of similar GC content and identical size, which indicates that hydrogen bonding and base stacking are conserved in the gas phase [17, 18]. Although similar relationships between solution and gas-phase stability were demonstrated for DNA and RNA duplexes, there is no comprehensive study that directly compares different types of nucleic acid duplexes.
The significance of the activation energy determined in MS/MS experiments is limited, when alternative fragmentation channels, such as nucleobase loss and backbone cleavage, compete with strand separation. High charge states promote strand separation  and favor asymmetric charge distribution within the duplex . The formation of backbone fragments, on the other hand, is promoted by high GC content  and generally indicates high gas-phase stability of the duplex. Double resonance experiments show that as in single strands, neutral base loss precedes backbone cleavage in the duplex . Gabelica et al.  described product ions formed after the ejection of backbone fragments from the intact duplex, which are believed to arise from backbone cleavage in the single stranded, fraying ends of the oligonucleotides. This result suggests that the strand separation occurs by gradual unzipping of the duplex rather than a two-state transition and gives direct evidence for parallel fragmentation reactions. Whenever multiple dissociation pathways are accessible, the preferred channel is determined by two factors, the activation entropy and the activation enthalpy . This reasoning was applied to duplexes , where changing the activation conditions was found to promote different dissociation channels. The authors argue that the rate constant for the strand dissociation increases more steeply with the internal energy of the system than the rate constant of the base loss. The entropy-favored channel can thus be accessed using fast activation regimes. Herein, we demonstrate how structural modifications can impact both the activation enthalpy and entropy of fragmentation channels and thus determine the preferred reaction pathway of nucleic acid heteroduplexes.
Oligonucleotides, Chemicals, and Solvents
The DNA decamer 5'-AACTGTCACG-3' was purchased from Microsynth (Balgach, Switzerland) and used without further purification. Two sequence isomers containing single tcDNA modifications were synthesized according to a previously published, adapted phosphoramidite chemistry procedure . The modified nucleotide in the decamer 10-2 (5'-AACTGTCACG-3') comprises an amide linker. The two tc-sugar-moieties in the decamer 10-3 (5'-AACTGTCACG-3') are fluorinated in the 2'-position. The oligonucleotides were diluted to a final concentration of 25 μM in a water:acetonitrile:TEA (49:49:2) solvent prior to analysis. Fluka HPLC water, Fluka HPLC acetonitrile, and triethylamine were purchased from Sigma-Aldrich (Sigma-Aldrich Chemie GmbH, Buchs, Switzerland).
For the characterization of duplexes, nine fully modified tcDNA oligonucleotides ranging from 10 to 15 nucleotides were synthesized. All fully modified tcDNA oligonucleotides contain thiophosphate linkers and a 5'-terminal phosphate group. Complementary DNA and RNA oligonucleotides were purchased from TriLink Biotechnologies (San Diego, CA, USA). They were used without further purification and redissolved in Fluka HPLC water to provide stock solutions of 1 mM. Equimolar amounts of complementary nucleic acids were combined in 250 mM NH4OAc to yield a final duplex concentration of 50 μM. A 7.5 M ammonium acetate solution purchased from Sigma-Aldrich was used to prepare the buffer solution for the annealing. The samples were annealed in a Eppendorf Mastercycler gradient PCR cycler (Vaudaux-Eppendorf AG, Schönenbuch, Switzerland). A temperature gradient was run from 90 °C to 18 °C over 2.5 h. For direct competition experiments, the tcDNA oligonucleotide was mixed at a 1:1:1 molar ratio with both complementary DNA and RNA strands in ammonium acetate buffer and annealed according to the same procedure. Prior to analysis, the duplex samples were diluted with an aqueous solution of 50% methanol to a final concentration of 2.5 μM. Fluka HPLC grade methanol (Sigma-Aldrich Chemie GmbH) was used to prepare the solvent.
All ESI-MS experiments were performed on a LTQ Orbitrap XL instrument (Thermo Fischer Scientific, Bremen, Germany) equipped with a nano-ESI source. The experiments were carried out in the negative ion mode with a spray voltage of –700 to –850 V. The capillary voltage was –45 V, the capillary temperature 150 °C, and the tube lens voltage was set to –200 V. Mass spectra were acquired in the FTMS mode from m/z 200 to 2000 with the mass resolution set to 100,000. Tandem mass spectrometric experiments were performed in the ion trap using helium as collision gas. The selection window for the precursor was defined as ±2.5 m/z. The activation time was set to 30 ms and the relative collision energies (rCE) applied for fragmentation ranged from 13 to 40. The Xcalibur Software Suite including Qualbrowser ver. 2.0.7 (ThermoFisher Scientific) was used for data processing. Data analysis was supported by the OMA and OPA software tool .
Complementary oligonucleotides were diluted in 20 mM ammonium acetate buffer solution to a final oligonucleotide concentration of 1 μM. Thermal melting curves were recorded in a Varian Cary-100 UV/VIS photospectrometer. The samples were heated to 85 °C; after that a cooling-heating-cooling cycle was run in the temperature range of 2–85 °C and a gradient of 0.5 °C per min. Three consecutive cycles were run and all annealing and melting curves were superimposed. The Cary WinUV software (Agilent Technologies, Santa Clara, CA, USA) was used for data acquisition. Data analysis was performed using the R software environment and the melting temperature was determined as the maximum of the first derivative of the superimposed curves.
Results and Discussion
Fragmentation Mechanism of tcDNA
Correlation of Solution and Gas-Phase Stability of DNA, RNA, and tcDNA Duplexes
Fragmentation Pathways of Sequence Identical Duplexes at a Charge State of –6. Parallel fragmentation channels are abbreviated as follows: D for strand separation, B for base loss and backbone cleavage, and T for the formation of truncated duplex ions. The backbone cleavage leading to truncated duplexes are indicated with red lines
Formation of Truncated Duplexes
For the two modified heteroduplexes, the collision energy was smaller than expected based on the stability determined by thermal melting. Moreover, the tcDNA duplexes stand out from the set of sequence-isomeric duplexes because an additional fragmentation channel was observed here. The loss of one or more nucleobases, mainly purines, from the intact duplex was detected. Nucleobase loss can induce the cleavage of the backbone to give raise to truncated duplex ions. For both heteroduplexes, D-tca2 and D-tcw3-AH ions were detected, which indicate backbone cleavage in the tcDNA strand after the 2nd or the 12th nucleotide, respectively. The formation of truncated duplexes was first reported for DNA duplexes in 2002 . The authors stated that truncated duplex ions can occur in precursors with relatively high gas-phase stability and are formed after partial unzipping of the duplex in the gas-phase. Our data demonstrate that the unwinding of the duplex takes place from both termini.
Truncated Duplex Ions Formed for tcDNA:DNA Heteroduplexes. The charge state was –6 for 14 and 15, and –5 for the shorter duplexes. The top sequence corresponds to the sugar-modified strand (C* = 5-methyl-cytosine), the complementary DNA strands are listed below. The listed percentage of the RNA heteroduplex was determined in direct competition experiments to assess the selectivity of tcDNA for the RNA versus DNA complements
For all tcDNA hybrids with 14 or 15 base pairs, the observed fragmentation channels include the release of nucleobases and backbone fragments as well as strand separation, hereafter referred to as base loss and dissociation, respectively. After dissociation of the hybrids, the intact DNA or RNA complements are detected, while the modified single strand further undergoes base loss and backbone cleavage. Backbone cleavage can occur before, during, or after full strand separation, which thwarts quantitative comparison of base loss and dissociation from the product ion spectra. Nevertheless, the release of up to four nucleobases from the duplex was observed in addition to the ejection of backbone fragments, which suggests that dissociation is only a minor fragmentation pathway of the tcDNA:RNA and tcDNA:DNA duplexes.
Parallel fragmentation channels in tcDNA:DNA heteroduplexes
The coincidence of two parallel reactions in tcDNA hybrids indicates similar values for the free activation enthalpies, ∆G‡, of both reactions. In the DNA:DNA duplex, by contrast, there is a significant difference in the free activation enthalpies for base loss and dissociation since only the latter is observed in CID experiments. We propose that the difference between the homo- and the heteroduplex is determined by two factors: First, the activation enthalpy for the base loss is reduced in the modified duplex. Second, the activation entropy for the dissociation is lowered because of the steric constraints of the tcDNA. Consequently, moving from the DNA:DNA duplex to the tcDNA hybrid, ∆G‡ of the base loss decreases and ∆G‡ of the dissociation increases, i.e., the free activation enthalpies of the two reactions converge. The low activation enthalpy for the base loss results from the destabilization of the N-glycosidic bond in the tc-sugar-moiety. As discussed for the singly modified DNA decamer 10-2, base loss and backbone cleavage are preferred at the modified site. The fact that backbone cleavage within the hybrid duplex only occurs in the tcDNA strand further evidences the weakness of the N-glycosidic bond in the modified nucleoside. As to the second effect, previously published thermodynamic data obtained from UV melting experiments  demonstrate that the formation entropy of tcDNA duplexes in solution is reduced relative to non-modified structures. Therefore, entropic stabilization of the hybrid duplexes should be assumed in the gas-phase.
As suggested by Gabelica et al. , the entropy-favored fragmentation pathway can be promoted by fast activation conditions. For the tcDNA hybrids, however, changing the activation time did not change the fragmentation pattern. Alternatively, the activation entropy can be studied indirectly when the activation enthalpy for the dissociation is altered. One approach is to compare different charge states since the Coulomb repulsion increases with the number of deprotonated phosphate groups and destabilizes the duplex. The tcDNA hybrids with z = –6 preferentially fragment by base loss. If the charge state of the precursor is increased to –7, strand separation becomes more pronounced, while base loss is reduced to the minor fragmentation pathway. A similar shift between competing reactions can be observed for the nonmodified DNA:RNA heteroduplex 15-1. Here, the formation of truncated duplex ions was observed for z = –5, whereas only dissociation is observed at higher charge states. While comparing charge states allows no fine-tuning of the activation enthalpy, it confirms that both reaction pathways, dissociation as well as base loss, are in principle accessible in modified and nonmodified duplexes.
Fragmentation Pathways of Mismatched tcDNA:DNA and DNA:DNA Duplexes at a Charge State of –6
Relative Stability of tcDNA:DNA and tcDNA:RNA Duplexes
The correlation between the collision energy and the melting temperature repeatedly reported within sets of similar duplexes does not extend to the comparison of different types of nucleic acid duplexes. In solution, the relative stabilities of six sequence-identical duplexes was found to increase in the order RNA:DNA ~ DNA:DNA < DNA:RNA < RNA:RNA ~ tcDNA:DNA << tcDNA:RNA. This trend does not correspond to the relative collision energies determined in MS/MS experiments. Moreover, dissociation, the favored fragmentation channel of nonmodified duplexes, constitutes only a minor pathway in tcDNA hybrids, whereas the formation of truncated duplexes is preferred. The free activation enthalpies of base loss and dissociation converge in tcDNA hybrids because the modified sugar-moiety decreases the activation enthalpy of the base loss and reduces the activation entropy of the strand dissociation. The decreased enthalpy for base loss is evidenced by the fact that bond scission preceding the ejection of backbone fragments from the intact duplex solely occurs from the modified strand owing to the weak N-glycosidic bond in tcDNA nucleosides. The activation entropy for the dissociation was studied by comparison of matched and mismatched duplexes. Herein, we report identical fragmentation patterns for two heteroduplexes based on the same tcDNA 15mer: the fully matched tcDNA:RNA hybrid and the tcDNA:DNA hybrid harboring two central C·T mismatches. This strongly contrasts with the difference in the melting temperature of 53 °C determined for the same two duplexes and illustrates the role of the activation entropy for strand separation in the gas phase. Consequently, tandem mass spectrometric experiments do not reflect the relative stability of duplexes when different types of nucleic acids are compared. For this reason, the stability of sequence identical tcDNA:RNA and tcDNA:DNA heteroduplexes is best assessed by direct competition experiments rather than CID. For the seven tested sequences, the tcDNA:RNA heteroduplex accounts for 66%–100% of the total signal intensity of all duplex species in equimolar mixtures of tcDNA and both DNA and RNA complements.
- 4.Pallan, P.S., Ittig, D., Heroux, A., Wawrzak, Z., Leumann, C.J., Egli, M.: Crystal structure of tricyclo-DNA: an unusual compensatory change of two adjacent backbone torsion angles. Chem. Commun. 7, 883–885 (2008)Google Scholar
- 5.Goyenvalle, A., Griffith, G., Babbs, A., Andaloussi, El, S., Ezzat, K., Avril, A., Dugovic, B., Chaussenot, R.E.M., Ferry, A., Voit, T., Amthor, H., Bühr, C., Schürch, S., Wood, M.J.A., Davies, K.E., Vaillend, C., Leumann, C., Garcia, L.: Functional correction in mouse models of muscular dystrophy using exon-skipping tricyclo-DNA oligomers. Nat. Med. 21, 1–9 (2015)Google Scholar
- 6.Aartsma-Rus, A., Van Deutekom, J.C.T., Fokkema, I.F., Van Ommen, G.-J.B., Den Dunnen, J.T.: Entries in the Leiden Duchenne muscular dystrophy mutation database: an overview of mutation types and paradoxical cases that confirm the reading-frame rule. Muscle Nerve 34, 135–144 (2006)CrossRefGoogle Scholar
- 8.Schürch, S.: Characterization of nucleic acids by tandem mass spectrometry—the second decade (2004–2013): from DNA to RNA and modified sequences. Mass Spectrom. Rev. (published online 2014) doi: 10.1002/mas.21442
- 12.Wang, B.H., Hopkins, C.E., Belenky, A.B., Cohen, A.S.: Sequencing of modified oligonucleotides using in-source fragmentation and delayed pulsed ion extraction matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Int. J. Mass Spectrom. 169, 331–350 (1997)CrossRefGoogle Scholar
- 23.Vekey, K.: Internal energy effects in mass spectrometry. J. Mass Spectrom. 31, 445–463 (1996)Google Scholar
- 25.Lesnik, E.A., Freier, S.M.: Relative thermodynamic stability of DNA, RNA, and DNA:RNA hybridduplexes: relationship with base composition and structure. Biochemistry 34, 10807–10815 (1995)Google Scholar