For years, mRNA circularization has been treated as a functional closed loop that helps translation. But single-molecule imaging reveals a striking possibility: translating mRNA can look compact without being physically connected at its ends.

Translation is mediated by precisely orchestrated sequential interactions among translation initiation components, mRNA, and ribosomes. Those interactions determine what occurs, and when, where, and in what order.

Most mRNAs are circularized via the eIF4E–eIF4G–PABP interaction, which stabilizes mRNAs and enhances translation by recycling ribosomes. The complication is that single-molecule fluorescence imaging has visualized complex data opposing the traditional functional circularization theory.

This walkthrough follows in vitro and live-cell imaging, then the questions raised by those results. Figure one introduces four single-molecule fluorescence approaches used in the mRNA circularization study. It shows how smFRET converts donor–acceptor distance into a fluorescence-transfer readout, smFISH uses multiple fluorescent DNA probes to locate individual mRNAs, and SiMPull captures complexes on a PEG-biotin surface for component-level analysis.

The final panel illustrates NCT or SINAPS, which distinguishes translating from non-translating mRNAs by labeling nascent peptides while MCP marks the mRNA itself. Single-molecule Förster resonance energy transfer, or smFRET, is one of the most widely used imaging techniques in this field.

It examines molecular interactions or the interdistance of residues by measuring nonradiative energy transferred from one fluorophore, the donor, to another, the acceptor. Because energy transfer occurs when the two fluorophores are within several nanometers, from one to ten nanometers, the strength and length of interactions can be determined in real time.

In purified systems, circularization is thought to begin when eIF4F—eIF4E, eIF4A, and eIF4G—recognizes the five-prime cap of mRNA as part of initiation control. Recognition of eIF4G increases eIF4E’s affinity for the cap. smFRET monitored fluorescently labeled eIF4E with and without eIF4G.

Full-length yeast eIF4G promoted an alternate RNP conformation that reached the ends of an mRNA. Pab1p, a yeast PABP, had an even more significant effect, causing both RNA ends to close even without poly-A or eIF4G. That suggests Pab1p can participate in RNA conformational changes through direct interaction with mRNA.

Figure two assembles single-molecule evidence for mRNA circularization in purified systems. It shows how eIF4F components alter cap binding, how mRNA structure and length affect eIF4E association, and how poly(A) RNA changes the bent conformation of PABP’s RRM domains, potentially strengthening eIF4G–PABP interaction and eighty-S formation.

Direct distance measurements place the investigated mRNAs at five to nine nanometers end to end, while tethering assays reveal heterogeneous monosome and polysome mRNP states. Translation is regulated by secondary structure, length, mRNA sequence, posttranscriptional modifications such as m6A modification, and poly-A length.

Even a small stem loop in the five-prime untranslated region was sufficient to perturb translation initiation and delay translation, in a way not detectable in an ensemble assay. The eIF4E–mRNA association is inversely proportional to the degree of secondary structure in the five-prime untranslated region.

Its association rate depends not on the free energy of that structure, but on its degree of complexity. The rate of association between eIF4E and mRNA tends to increase as total mRNA length shortens, while eIF4G–mRNA interaction accelerates eIF4E association in proportion to mRNA length.

Conformational changes in a regulator can also play a crucial role in circularization. When PABP binds a poly-A tail, conformational changes in its RRM2–RRM3 region affect PABP–poly-A and PABP–eIF4G interactions. Mutation of that region inhibits the PABP–eIF4G interaction and efficient formation of eighty-S ribosomes.

The result supports the importance of PABP binding to the poly-A tail in bringing mRNA ends into proximity. In cells, measuring the distance of the mRNA compartment is described as the most direct, although primitive, way to investigate circularization.

Morisaki and colleagues measured distances between nascent peptide chains and the three-prime untranslated region for polysomes with lengths of one hundred twenty-five, three hundred seventy-four, and one thousand five hundred forty-four amino acids. Those experiments found distances of sixty-five to one hundred five nanometers, shorter than expected, suggesting that polysomes are compact rather than extended.

Yet the size was not correlated with mRNA length in that experiment. More recent work reported that a closed loop is not a stable state of translating mRNA. smFISH showed that mRNP compaction depends on translation state, with ribosome release as the main cause of compaction when initiation is inhibited.

Figure three shows how single-RNA imaging measures mRNA circularization in cells. Panel a uses the distance between nascent peptide epitopes and a three-prime UTR marker to estimate the compactness of translating mRNA, while panel b measures both ends directly with smFISH and shows end-to-end distance increasing with ribosome occupancy.

Panel c compares cap-driven, IRES-driven, and combined translation, illustrating that translation state and ribosome loading shape mRNA compaction. The cellular result is not limited to eIF4E–eIF4G–PABP. In an IRES-mediated study, a reporter mRNA contained a cap and an IRES downstream of the cap open reading frame, producing a ten-times FLAG epitope and a twenty-four-times SunTag.

The distances from the cap open reading frame and IRES open reading frame to the three-prime untranslated region were measured during cap-only, IRES-only, or cap-plus-IRES translation. The IRES open reading frame was compact when idle, possibly because it lacked ribosomes.

In both cases, the open-reading-frame to three-prime-untranslated-region distance was proportional to intensity, meaning the number of ribosomes. With cap-plus-IRES translation, the IRES distance decreased significantly, consistent with fewer available ribosomes and fewer ribosomes loaded on the mRNA.

These observations create a direct tension between single-molecule and ensemble views. Biochemically, circularization should depend on protein–protein and protein–mRNA binding affinities and on protein concentrations in cells, while translation elongation and initiation occur during the same process.

The review therefore asks how ribosomes are reinitiated after termination, whether communication between both mRNA ends occurs only in a nontranslating state, and whether loaded ribosomes can break eIF4G–PABP or eIF4E–eIF4G interactions. One possibility is that functional circularization occurs after the first ribosome completes translation: the first ribosome may drag eIF4F to meet PABP at the three-prime end and make the physical link.

A complete account must also address circularization mediated by structural IRESs, m6A, and three-prime cap-independent translational enhancers, alongside classical five-prime-cap and poly-A mechanisms. The smFI methods may help answer these questions. But limitations remain.

It is still difficult to track single or multiple translation regulators simultaneously because of the high intracellular concentration of the targets, and further advances in protein-labeling strategies are required. Functional mRNA circularization based on protein–protein or RNA–protein interactions has been considered a canonical dogma in translational studies.

Recent single-molecule studies, however, show that mRNA can appear circularized yet cannot be physically connected by proteins during active translation. smFI does not provide a complete solution, but it is currently the most optimal method for visualizing the dynamic behavior of a single mRNP in intact cells.

Its future value is a more comprehensive understanding of how interactions in an mRNP complex change, and how the resulting mRNA conformation changes throughout translation. The central takeaway is that active translation and mRNA circularization are not simply the same state.

Single-molecule fluorescence imaging shows dynamic, compact mRNPs and exposes questions that ensemble measurements cannot resolve.