Characterizing RNA Molecules with FTIR
Although vibrational spectroscopy has long been used in structural analysis of DNA, the field has only recently expanded to address ribonucleic acid
Methods of vibrational spectroscopy are perfectly suited to the measurement and characterization of nucleic acids. Through derivation of peaks from unique absorption spectra, investigators can identify and visualize variability in intra- and intermolecular interactions, as well as association with pharmacological modulators and biopolymers, including proteins. Analyses, modeling, and prediction of regulatory processes and guided design of therapeutic agents can advance rapidly via these and other types of studies. Although vibrational spectroscopy has long been used in structural analysis of DNA, the field has only recently expanded to address its more ephemeral relation, ribonucleic acid (RNA). This is both fortuitous and a natural reflection of our current era, as appreciation for the diversity and topographical charisma of RNA molecules has blossomed while laboratories continue to reveal the myriad roles of non-coding RNAs.
Distinct from DNA, RNA has the property of being single-stranded, imparting to it a sticky nature in which kinks and hairpins arise, and in which base pairings can easily diverge from the canonical Watson-Crick couplets. Bases commonly engage in guanine-uracil wobbles and Hoogsteen arrangements, while also forging cross-strand hydrogen bonds, resolving into secondary and tertiary structures. These chemical and architectural variations often precipitate changes in biological and regulatory function. They can contribute to determining, for instance, the efficiency with which a short RNA oligonucleotide silences target gene expression, or the rate at which transfer and ribosomal RNAs can control protein synthesis. However, RNA molecules themselves are labile and refractory to crystallization, necessitating characterization strategies based on indirect means.
FTIR: principles, and RNA by the numbers
Infrared (IR) spectroscopy operates as a variation on the principle of the Michelson interferometer, in which a light source is projected through a beam splitter with the addition of a movable mirror to disperse the infrared spectrum, and a detector, usually composed of triglycine sulfate (DTGS), records absorption. Absorption bands generated by a sample comprise an interferogram, which can then be converted to peaks corresponding to wavenumbers by using the appropriate Fourier transform function, a mathematical manipulation that allows the expression of time- and space-dependent events in terms of frequency.
For RNA, the relevant wavenumbers occupy the range of about 2,000 to 700 cm-1, in which there are four defined spectral domains: 1) out-of-plane base vibrations (800-760 cm-1); 2) in-plane base vibrations (1750-1450 cm-1); 3) sugar conformations (865-820 cm-1,); and 4) phosphate conformations (1245-1220 cm-1). Within these ranges, base, sugar, and phosphate vibrations can be detected with a high degree of sensitivity and specificity, with positions and interactions precisely inferred from the derived spectra. For instance, an absorption band at 781 cm-1 corresponds to an out-of-plane guanine nucleotide normally bound to a cytosine, whereas a band at 778 cm-1 denotes a noncanonical guanine-uracil wobble pairing.
FTIR setups: Transmission and reflection, and their comparative advantages
There are two commonly utilized setups for the FTIR study of RNA: transmission and reflection. In a transmission setup, the sample directly absorbs IR light and must be in an IR-transparent cell, usually composed of zinc selenide (ZnSe). Calcium fluoride is more pressure-resistant, and it is also employed in transmission setups. However, its transmission range is inadequate to cover wavenumbers below about 1,000 cm-1 so it can be problematic for RNA structural measurements.
In a typical FTIR run, an RNA sample can be generated by either in vitro transcription or oligonucleotide synthesis, and then heated and cooled to allow native secondary structure to form. It is deposited onto the ZnSe surface, dried into a thin film by applied air or under gas flow, and the instrument commences. After obtaining an interferogram, an investigator generates spectra by subtracting baseline and solvent readings, obtaining the second derivative of absorption bands, and employing Lorentzian and Gaussian curve fit algorithms. These manipulations can now be seamlessly performed in real time or at end points, with cursory training, using dedicated software.
There are a couple of caveats to using transmission FTIR to study RNA. One is that ordinary deionized water (H2O) confounds acquisition because it displays high absorbance at 1,650 cm-1, in the range of base vibrations; consequently, investigators using transmission instrumentation substitute deuterated “heavy” water (D2O). Secondly, because of the high degree of sensitivity, baseline and sample measurements must ideally be taken by the same user using the same cell, and in a short time frame, because even small variables such as emitted heat from optics can impact the consistency of absorption readings. Moreover, when working with RNA, all the usual precautions apply, such as using RNAse-free solutions, and additional purification steps to remove enzymes and contaminating nucleic acids from reaction products.
For reflection studies on the other hand, investigators typically use an attenuated total reflection (ATR) setup, in which an IR beam reflects one to several times through a laterally positioned crystal of either ZnSe or diamond to create an evanescent wave. Some of the resulting energy is absorbed by the sample and the radiation returned to the detector, from which analysis proceeds. This setup is advantageous in that H2O can be used as a solvent, and liquid samples can be simply dropped onto the crystal stage, rather than requiring the drying step. However, penetration of the reflecting crystal is wavelength-dependent, and resulting spectra must be corrected to conform with standard analysis of analogous transmission spectra.
As a practical example, investigators have characterized associations of nucleic acids with a class of chemotherapeutic agents called anthracyclines, which halt cancer cell proliferation by intercalating DNA. Additional interactions between drug family members and RNA molecules may vary and have the potential to impact both mechanism and function.
Doxorubicin and epirubicin are both approved anthracycline chemotherapeutic agents. Epirubicin is a more recent structural analog of doxorubicin, epimerized at the C-4 hydroxyl group. Doxorubicin is broadly used, but exhibits cardiotoxicity in some patients, and epirubicin demonstrates more favorable pharmacokinetics and reduced toxicity, meaning it can be delivered at higher doses. It has been shown to associate with, but not modify, guanine and uracil bases on tRNA molecules, suggesting that although it functions by intercalating DNA, its effects on key regulatory RNAs may be negligible. Within this and other classes of pharmaceuticals, better understanding of drug-nucleic acid interactions can potentiate medicinal chemistry strategies to improve efficacy and decrease toxicity through off-target effects. Extending this principle, non-coding RNAs can potentially serve as therapeutic targets, and new classes of RNA-binding antiviral and anti-cancer drugs can be developed, with FTIR serving as a powerful analytical tool.
References:
1. Wien et al (2021). SRCD and FTIR spectroscopies to monitor protein-induced nucleic acid remodeling. In: Boudvillain M (eds) RNA Remodeling Proteins. Methods in Molecular Biology, vol 2209. Humana, New York, NY. https://doi.org/10.1007/978-1-0716-0935-4_6