CCR5 frameshift signal questioned by retraction
A Signal That Wasn’t There
In July 2014, a research team including Jonathan D. Dinman at the University of Maryland published a study describing a programmed -1 ribosomal frameshift signal in the human CCR5 mRNA — the messenger RNA encoding the co-receptor that HIV-1 uses to enter cells — that appeared to make ribosomes slip back one base. [1] The finding carried weight because it suggested a cellular gene, not just a virus, could be regulated this way.
A second line of work, published in 2022 by Y. A. Kahn and colleagues, approached the same CCR5 sequence from a different direction. Their question was simpler and more skeptical: does the frameshift signal actually fire, or does the assay merely look like it does? Where the original study read a signal, the replication effort read an artifact. Two approaches, one sequence, two incompatible conclusions — and the resolution would come not from argument but from the design of the measuring instrument itself. [1].
The instrument in question was a dual luciferase reporter vector called p2luci1. In that construct, transcription is driven from an SV40 promoter and is enhanced by the presence of an artificial intron. The CCR5-derived sequence was cloned into this vector, and the readout seemed to show frameshifting. The catch, uncovered only through further investigation, was that the CCR5-derived sequence contained a cryptic 3’ splice acceptor site, resulting in the generation of artefactual splice variants. [1].
A Cleaner Instrument, a Different Answer
To test their own result, the authors turned to an improved version of the dual luciferase reporter. In this newer construct, the SV40 promoter was replaced by the strong CMV promoter, and the artificial intron was absent. Same sequence, cleaner instrument. The CCR5-derived sequence did not promote detectable levels of programmed -1 ribosomal frameshifting. [1] The paper’s central claim — that CCR5 mRNA carries a functional -1 PRF signal — no longer has data behind it that the editors trust. These findings have been replicated by at least two other groups, suggesting that the perceived frameshifting activity was due to a splicing artefact. [1].
Ashton Trey Belew, Sharmishtha Musalgaonkar, Vivek M. Advani and Sergey O. Sulima agree with the retraction. [1] Arturas Meskauskas, Wojciech K. Kasprzak, Bruce A. Shapiro and Jonathan D. Dinman disagree. Four against four, with the disagreement preserved in the public record rather than smoothed over.
