Depurination of DNA and RNA sequences is an alteration of the sugar-phosphate backbone in which hydrolysis of a purine base (adenine or guanine) occurs. Synthetic oligonucleotides suffer from such errors during the assembly process, resulting in unwanted and harmful contaminants. Contemporary DNA and RNA synthesis utilizes solid-phase phosphoramidite chemistry to construct nucleotide sequences through a succession of phosphodiester linkages.
By incorporating dimethoxytrityl (DMT), an acid labile 5´ protecting group, nucleotide assembly is accomplished through sequential automated cycles of deprotection, coupling, capping, and oxidation. The greatest potential for damage to the sequence occurs during deprotection (detritylation) where the oligonucleotide is exposed to dilute acid, increasing the risk of purine hydrolysis. Adding further complexity, the depurination rate of single-stranded DNA (as synthetically produced) is fourfold greater than native double-stranded DNA.
Unlike a living cell, the synthetic process lacks the necessary mechanism to selectively clear apurinic sites from the full-length sequence. Consequently, to deliver a pure and error-free product, optimized synthesis and purification techniques are required to minimize and remove depurinated fragments and other remnant impurities.
Crude synthetic oligonucleotides are purified by either trityl-on or trityl-off methodologies. The trityl-on method offers several key advantages, the most attractive being that it involves fewer steps and is more compatible with automation and high-throughput applications.
In trityl-on purification protocols, the final 5´ DMT protecting group is retained on the nucleotide and removed during purification. In trityl-off applications, the 5´ protecting group is cleaved in the oligonucleotide synthesizer prior to purification. An advantageous feature of trityl-on separation is the lipophilic properties of the DMT group, which can serve as an idyllic handle to enable discrimination between the protected full-length sequence and synthetic contaminants.
Trityl-on techniques are preferred for high-throughput parallel processes because they do not require chromatographic instrumentation. But a drawback has been the potential for post-purification depurination.
During synthesis, the integrity of the ring structures of the purine bases is secured and shielded from acid hydrolysis with benzoyl and isobutyryl protecting groups. These groups are removed soon after support release, however, increasing the likelihood of nucleobase cleavage when exposed to an acidic environment. Consequently, trityl-on techniques require a balance of avoiding purine hydrolysis while ensuring complete detritylation.
The drawbacks of this method have limited the acceptance of trityl-on techniques, despite their advantages. Since their commercial introduction, cartridge-based and high-throughput trityl-on products have delivered varying levels of depurination, according to users. Some have mentioned finding little or no depurination while others have observed considerable purine hydrolysis.
The purpose of the study described in this article was to better gauge the factors influencing purine hydrolysis during trityl-on practices using an appropriate array of sequences and conditions. In this investigation we used Clarity QSP from Phenomenex (www.phenomenex.com), a high-throughput, cartridge-base trityl-on purification product.
Crude DNA sequences were purified via the Clarity QSP trityl-on cartridge product using a standard protocol with aqueous dichloroacetic and trichloroacetic acid used for detritylation. Final eluting solutions were varied utilizing either acetonitrile-water or a Na2CO3-buffered acetonitrile solution.
To more accurately measure depurination, DNA sequences were constructed with lone internal purine bases as well as with terminal and internal purine bases. In addition, pyrimidine compositions were varied among the sequences to examine possible neighboring influences. Polyamine hydrolysis was incorporated with anion exchange chromatography to qualify and quantitate the level of depurination following Clarity QSP purification.
Aliquots of the final elution from the QSP cartridge were dried using a N2 purge. Each pellet was reconstituted in 50 mM Hepes/2mM Spermine/1 mM EDTA (pH 8.1) and incubated at 37°C for one hour. The hydrolysis mixture was diluted 10-fold prior to loading on an IEX column and analyzed. Quantitative values of depurination were determined based on comparing degraded and intact peak areas. Electrospray mass spectrometry was utilized in conjunction with the QSP automated 96-well format to monitor post-purification depurination of longer DNA sequences with high purine content.
Results and Discussion
Findings disclose that eluting conditions had a far greater influence on depurination than sequence composition or deblocking conditions. Somewhat surprising was the minimal level of depurination occurring during detritylation. Even several hours following column elution, little depurination of the oligonucleotide had occurred regardless of elution solvent. After ambient overnight incubations, however, results varied considerably as depurination was observed to have drastically increased in nonbuffered pH elutions (water–acetonitrile) using 3% DCA for detritylation (Figure 1).
In stark contrast, sequences eluted in a pH-buffered solvent showed virtually no depurination even when 15% TCA was used for deblocking (Figure 2). The rate of depurination was witnessed to have accelerated among those sequences with a terminal purine when compared to lone internal purine base sequences.
Prior research has observed similar findings, and we postulate that this is the result of the 5´ phosphate group at the apurinic site serving as an electron-withdrawing group further enhancing chain cleavage. When pyrimidine groups were placed at the 5´ terminal, depurination rates diminished considerably. While guanine is reported to release at a slightly faster rate than adenine, our observations yielded no significance difference between the purine bases among the tested sequences. Moreover, adjacent pyrimidine groups were found to have no prevailing influence on depurination.
Today, the majority of synthetic oligonucleotides are produced in combinatorial-style multiplex formats enabling oligo manufacturers to produce tens of thousands of sequences per day. To meet their purification demands, manufacturers are relying on parallel platforms to work in concert with automated liquid-handling systems.
The Clarity QSP product is offered in a 96-well plate that is designed specifically for robotic systems. High-throughput processes can augment depurination as acid exposure times are often longer during detritylation than with cartridge formats. Apurinic occurrence, however, can be avoided simply by using a lower acid concentration during detritylation.
For instance, Figure 3 presents an ESI-MS analysis that shows reducing acid strength to 1% DCA and eluting in a physiological pH buffer averts any depurination and without compromising complete trityl release.
The gathered data presents clear evidence that the primary source for post-column depurination is the pH of the eluting solvent, whereas deblocking conditions and sequence composition are rather minor contributors toward depurination. Further, results clearly show that simple acetonitrile/water mixtures are insufficient to neutralize the eluted oligonucleotide; some buffering to neutral pH is required to prevent depurination.
Our investigation served to address the causes behind the conflicting opinions about trityl-on techniques and depurination. By reducing acid concentration, limiting exposure times, and utilizing physiologically accommodating elution buffers, we were able to substantially minimize, if not eliminate, depurination from occurring during and after on-column detritylation. In addition, sequence composition, considered by many as a primary cause for purine hydrolysis, was discovered to be of little influence under appropriate conditions.
Trityl-on purification has its fair share of opponents and proponents, and this study was designed to evaluate concerns regarding trityl-on purification and its assumed proliferation of depurination. These results clearly show that with a proper protocol, trityl-on purification does not contribute to the degree of depurination of a synthetic oligonucleotide.
Gregory Scott (firstname.lastname@example.org) is research scientist, Cathy Cordova is brand manager, and Michael McGinley is product manager at Phenomenex. Web: www.phenomenex.com.