What are the reasons for using oligo-dT instead of oligo-U to isolate mRNAs?

Question

dTTP oligonucleotides are used to isolate mRNAs because mRNAs (in eukaryotes) have a poly A tail which binds to the complementary oligo-dT. However, why do we not use oligo-U instead (uracil)? I would guess it is because ribonucleotides have the additional 2' OH which is reactive, so less desirable, but is there any other reason?

Answer 1

Note - This answer only covers the comparison of rA-dT (RNA:DNA) and rA-rU (RNA:RNA) duplexes, and not more exotic options involving deoxyuridine (dU) and ribothymidine (m⁵U).

I can think of three possible reasons why oligo-dT may be preferred over oligo-rU for poly(A) hybridization and mRNA capture:

1. The rA-dT duplex is more stable than the rA-rU duplex, thus affording a more efficient pulldown.
2. dT oligonucleotides are easier to synthesize or more stable than rU oligonucleotides (and therefore cheaper to manufacture).
3. There are methodological considerations beyond the hybridization step that make oligo-dT a better choice.

Let's address each of these...

---

Hypothesis #1 $\rightarrow$ The rA-dT duplex more stable than the rA-rU duplex.

Conclusion #1 $\rightarrow$ probably not

Using NMR to observe the exchange rates of imino protons, Huang and Russu compared the stabilities of homologous duplexes containing tracts of rA-dT, dA-rU, rA-rU, and dA-dT base pairs.¹ Overall, they found the rA-dT and rA-rU duplexes to be similar in terms of base-to-base stability:

(i) Among the four nucleic acid structures investigated, the highest base-pair stabilities are observed for the dA-dT tract of the DNA duplex; the lowest base-pair stabilities are observed for the dA-rU tract of the dA-rU RNA-DNA hybrid.
(ii) The stabilities of individual base pairs in the rA-dT tract of the rA-dT hybrid are the same as those of rA-rU base pairs in the RNA duplex and are intermediate between the stabilities of DNA dA-dT and hybrid dA-rU base pairs.
(iii) The stabilities of base pairs in the rA-dT tract of the rAdT hybrid and in the rA-rU tract of the RNA duplex do not vary very much with the location of the base pair in the tract. In contrast, in the DNA duplex and the dA-rU hybrid, the highest and lowest stabilities are observed for the central base pairs in the tract.

This opposes an argument that oligo-dT is preferred due to enhanced stability of the resulting RNA:DNA duplex. An important note here is that the study cited above only tested tracts between 4 and 10 base pairs long, whereas most mRNA capture protocols utilize dT oligos in the 12-25 nt range.²

---

Hypothesis #2 $\rightarrow$ dT oligonucleotides are easier to synthesize or more stable than rU oligonucleotides.

Conclusion #2 $\rightarrow$ yes

As stated in the question, RNA is generally considered less stable than DNA due to the reactivity of the 2′-hydroxyl group. During solid-phase RNA synthesis, iterative protection and deprotection of the 2′-OH is required for addition of each new nucleoside, meaning that there are many more steps in RNA synthesis than solid-phase DNA sythesis when comparing equal-length oligos.³ Practically, this means rU oligonucleotides are much more expensive to order that dT oligonucleotides. As of 12 July 2021, IDT quotes me \$8.55 for 100 nmol oligo(dT)₁₉, whereas the same amount of oligo(rU)₁₉ costs \$142.50.

Beyond the cost, the reactivity of RNA and the ubiquity of RNases means one must take special precautions when handling and storing of RNA oligos, as explained by IDT:

The inherent chemical structure of RNA makes it less stable than DNA because the additional hydroxyl group attached to the 2’ position of the pentose ring in RNA nucleotides makes RNA more liable to alkaline hydrolysis. In addition, RNases are pervasive, making RNA more susceptible than DNA to degradation. Thus, RNA oligonucleotides should be stored as an ethanol precipitate at –80°C for greater stability. For short-term storage, we recommend resuspending RNA oligos in a nuclease-free neutral to slightly acidic buffer containing a chelating agent, such as IDTE. However, it is always important to avoid RNases and alkaline hydrolysis when RNA oligos are stored.

---

Hypothesis #3 $\rightarrow$ There are methodological considerations beyond the hybridization step that make oligo-dT a better choice.

Conclusion #3 $\rightarrow$ yes

In state-of-the-art mRNA library preparation kits such as those available from New England Biolabs and ThermoFisher Scientific, the bead-bound dT oligos serve both as the baits for poly(A) hybridization and as the primers for first-strand cDNA synthesis. These kits use RNA-directed DNA polymerase (e.g. MMLV Reverse Transcriptase) for cDNA synthesis, which requires a DNA primer from which to polymerize new DNA. Thus, any protocol utilizing an RNA:RNA hybridization at the pulldown step would have to dissociate the strands and introduce DNA primers before reverse transcription.

---

References

1. Huang Y, Russu IM. Dynamic and Energetic Signatures of Adenine Tracts in a rA-dT RNA-DNA Hybrid and in Homologous RNA-DNA, RNA-RNA, and DNA-DNA Double Helices. Biochemistry. 2017 May 16;56(19):2446-2454.
2. Anderson AJ , Culver HR , Prieto TR , Martinez PJ , Sinha J , Bryant SJ , Bowman CN . Messenger RNA enrichment using synthetic oligo(T) click nucleic acids. Chem Commun (Camb). 2020 Nov 21;56(90):13987-13990.
3. RNA Oligonucleotide Synthesis. ATDBio.