2.1. Isolation of tRNA from Biological Samples

There are several methods reported for the isolation of tRNA adjusted to laboratory budget and scale. Thus, while methodologies vary between laboratories, the underlying principles remain constant. First, cells are lysed or homogenized using TRIzol (phenol, chloroform, guanidinium thiocyanate) with agitation, or mechanically, using a bead beater or French pressure cell in presence or absence of lysozyme. Denaturants such as phenol and chloroform, and chaotropic agents, like guanidinium thiocyanate, also quench biochemical processes during isolation, thereby maintaining RNA integrity. Commonly, cells are suspended in aqueous buffer prior to homogenization and subjected to separation between aqueous and organic phases. The use of acidic phenol (pH ~4.5) enables retention of RNA in the aqueous phase, while DNA is extracted into the organic phase. Once RNA is sufficiently isolated into the aqueous phase, it can be further purified using lithium chloride treatment which precipitates large RNA fragments, but not tRNA or small RNA fragments [115]. The resulting soluble material can be precipitated with isopropanol or ethanol, or retrieved with silica-based spin columns. While alcohol precipitation is certainly more cost-effective, yields are lower, and caution must be taken not to over or not sufficiently dry the resulting RNA pellet. Thus, although the RNA isolation process can be flexible, caution is advised to ensure efficient isolation of pure, high quality tRNA samples. This can be evaluated using UV absorbance, denaturing gel electrophoresis, or fluorescent dye-based electrophoresis [116].

Many studies require the purification of individual tRNA species, and advances in liquid chromatography (LC) techniques have streamlined this process as opposed to the cumbersome traditional methods involving gel electrophoresis. There are several chromatography techniques facilitating RNA purification including anion-exchange and ion-pair reverse phase chromatography [116]. Affinity-based approaches for purification of specific tRNA species make use of oligonucleotide probes designed to capture the tRNA isoacceptor of interest using a unique sequence complementary to that within the tRNA [117,118,119]. Though the presence and locations of modifications within tRNA species are not fully known across organisms, the genomic sequences encoding tRNAs are commonly available, and thus can be used for probe design. Affinity purification methods make it possible to isolate bulk amounts of a specific tRNA isoacceptor, which is particularly useful when assessing modification function by comparing structure and/or function of synthetic hypomodified tRNA to biologically isolated and thus, fully modified tRNA species.

In vitro transcription methods to generate synthetic tRNA transcripts enable efficient and robust production of a pure single tRNA isoacceptor. However, the products are completely devoid of modifications, and, in some cases, subsequent kinetic analysis of these transcripts can skew assay results away from physiological relevance. Recent reports have demonstrated that modifications not only modulate the structure and function of these adapter molecules, they also decorate tRNA molecules to mediate specificity in substrate recognition by their enzyme interacting partners. One example is that of a study in which Rodriguez-Hernandez et al. conducted an exhaustive structural and functional investigation of the role of sulfur in tRNA [6]. Aminoacylation and ribosome utilization in Escherichia coli was assessed when using synthetic unmodified tRNA transcripts and fully modified tRNA obtained via affinity purification. Structural work showed that hypermodified tRNA^Gln induces conformational changes in glutaminyl-tRNA synthetase (GlnRS) which improve the conformation of a surface loop within the protein, and create a specific binding pocket for the 2-thio moiety. Aminoacylation kinetic analysis demonstrated that E. coli GlnRS had a 10-fold improvement in binding affinity for tRNA^Gln containing s²U compared to unmodified tRNA^Gln, which was further improved five-fold with the addition of the cmnm⁵-modification. Furthermore, thiolated U34 tRNA improved binding affinity to Gln codons and demonstrated five-fold enhancement of GTP hydrolysis by E. coli EF-Tu (elongation factor thermo unstable) compared to the unmodified tRNA [6]. Together, these findings provide excellent examples of the effects of tRNA modifications on enzyme kinetics and highlight the importance of comparing results from methods using both synthetic and native tRNA molecules.

Historically, chemical reagents have been used to selectively label modified nucleosides based on the chemical functionality of such modifications [120]. With the rise of mass spectrometry and other sensitive techniques, chemical labeling and modification conversion techniques remain useful as they provide improved detection and chemical stability for target analytes. Thiol groups are strong nucleophiles able to react with a wide range of reagents, including halo-acetamides (s²U, s⁴U) [121], 4-bromomethyl-7-methoxy-2-oxo-2H-benzopyran (bromomethylcoumarin) [122], and 3-carboxy-2,2,5,5-tetramethyl pyrroline-1-oxyl anhydride with ethyl hydrogen carbonate (mnm⁵s²U) [121]. Labeling and detection of other non-thiolated modifications has also been described in the literature. For example, isothiocyanate and activated amines (e.g., ethylenediamine) can covalently interact with the carboxyl group on t⁶A [123] and aliphatic amines can attack queuosine [124]. Although these reagents have specific selectivity towards target modifications, other modifications sharing similar functional groups can display cross-reactivity. For instance, in addition to bromomethylcoumarin’s ability to target thiolated uridine [122], it also reacts with pseudouridine [125], and potentially with uridine and thymidine [126].

Thiophilic “soft” metal mercury reacts readily with sulfur-modified nucleic acids. Igloi and his colleagues in 1988 first introduced a synthetic [(N-acryloylamino)phenyl]mercuric chloride (APM), which can be co-polymerized into polyacrylamide gels to separate thiolated tRNA [127]. APM was further developed by the Biondi group into a three-layer polyacrylamide gel with only the middle layer containing a high amount of the organomercuric compound [128]. The formation of a coordinate covalent bond between Hg and S ligand retards the electrophoretic migration of thio-modified RNA. The results can be visualized by standard methods for polyacrylamide electrophoresis imaging, including fluorescence, silver or ethidium bromide staining. Although the APM-gel has proved to be a low-cost and effective detection method, different thiol-containing tRNAs have shown different migration patterns through the APM layer due to varying structures of the thionucleotide. The APM-gel up to date has shown a great ability to covalently link 4-thiouridine and 2-thiouridine and its derivatives. However, when the sulfur is not in the thiocarbonyl form, for instance, in ms²i⁶A or ms²t⁶A modifications, APM has not been found to covalently coordinate with these thio-modified nucleosides effectively [129]. Despite its low selectively towards different sulfur modifications in tRNA, APM gel separation is still one of the most commonly used detection and isolation methods for thiolated RNA.

Queuosine is a non-thiolated modification, but its biosynthesis depends on [Fe-S] clusters (Table 1). This modification can also be retained in a polyacrylamide gel containing a boronic acid derivative. The synthetic N-acryloyl-3-aminophenylboronic acid (APB), when co-polymerized into acrylamide gel, is able to retain cis-diol groups present in tRNA, like queuosine, resulting in separation of the cis-diol containing ribonucleic acid [130,131]. However, APB is not only able to separate queuosine, but any tRNA samples containing cis-diol, including RNA with non-phosphorylated 3′ ends [130] and the newly discovered bacterial nicotinamide adenine dinucleotide (NAD)-capped tRNA [132]. From both APM- and APB-gel detection, thionucleosides may be extracted and isolated for further quantification.

Northern blot has been a valuable tool for probing individual tRNA species. Sequence specific hybridizing probes are usually modified with either radioisotope ³²P or fluorescent labels to improve detection sensitivity. Although radioactive-labeled probes are advantageous for improving detection sensitivity and assay quality, the disadvantage of such experiments is the need to minimize the usage of radioactive substances. This technique has also been used in conjunction with APM-containing polyacrylamide gels to identify the occurrence of thiolation on certain tRNA species [133,134,135,136,137]. Recently, a variation of the standard method, immuno-Northern blot, was provided which uses antibodies that specifically bind with modified nucleosides such as 1-methyladenosine (m¹a), N⁶-methyladenosine (m⁶A), pseudouridine, and 5-methylcytidine (m⁵C) [138]. This highly sensitive and relatively simple protocol enables small laboratories to compare the abundance of modified nucleic acids across samples.