Less is more

Pan, Meng, Xiaorong Yang, Lei Zhou, Xinna Ge, Xin Guo, Jinhua Liu, Dabing Zhang, and Hanchun Yang. 2012. “Duck Hepatitis a Virus Possesses a Distinct Type IV Internal Ribosome Entry Site Element of Picornavirus..” Journal of Virology 86 (2): 1129–1144.

In this study the authors have characterized IRES found in 5'UTR of DAHV. Almost all the data presented here are in line with earlier suggestion that this IRES belongs to wide group of hepacivirus/pestivirus (HP) IRESs (1). Not in line with this suggestion is the finding that translation of a bicistronic mRNA with DAHV IRES in the intercistronic position is highly sensitive to eIF4G cleavage by SVDV 2A protease.

However, this test was performed in vivo by means of DNA transfection. This approach suffers from a plenty of possible artifacts, for example cryptic promotor activity, e.g. present in the HCV IRES cDNA (2). One might argue that using T7 RNA polymerase vaccinia virus system overcomes these shortcomings, but the transfected DNA goes to the nucleus anyway and can be transcribed there, generating short capped monocistronic Fluc containing mRNAs, which translation would be perfectly inhibited by 2A protease. Therefore, RNA controls should have been performed: Nothern, RT-PCR, or RNAi against the first cistron (3). Also, a simple way to show that the intact eIF4G is required is translation in vitro with the addition of recombinant protease.

Funny is the authors' statement about "~0.5-fold stimulation" (and it was stimulated indeed) of the EMCV IRES which normally means 2-fold inhibition. :) Finally, the authors draw their conclusion:

These data indicate that the intact eIF4G is required for DHAV internal initiation of translation and that the HAV IRES element is no longer the only one abolished by cleavage of eIF4G.

Well, not exactly. Translation of the DAHV IRES was stimulated ~0.4-fold inhibited ~2.5 fold upon eIF4G cleavage. (Where is a control with HCV IRES, by the way?) This is not what one would call abolished. If a factor is required, there's no translation in it's absence: EMCV IRES translation is abolished by eIF4A R362Q mutant. Therefore, you can't contend that translation is eIF4G-dependent until you perform a set of in vitro translations with addition of hippuristanol (4) or eIF4A R36Q mutant or anything of this kind.

That's surprising, because this fact is the only one in the whole manuscript that makes a researcher blink at. But there is a single experiment lacking controls. And you can be sure that in a forthcoming review this IRES will be cited as a novel. Which is not proved yet. 

1. Hellen, Christopher U T, and Sylvain de Breyne. 2007. “A Distinct Group of Hepacivirus/Pestivirus-Like Internal Ribosomal Entry Sites in Members of Diverse Picornavirus Genera: Evidence for Modular Exchange of Functional Noncoding RNA Elements by Recombination..” Journal of Virology 8 (11): 5850–5863.

2. Dumas, Estelle, Cathy Staedel, Marie Colombat, Sandrine Reigadas, Sandrine Chabas, Thérèse Astier-Gin, Annie Cahour, Simon Litvak, and Michel Ventura. 2003. “A Promoter Activity Is Present in the DNA Sequence Corresponding to the Hepatitis C Virus 5' UTR..” Nucleic Acids Research 3 (4): 1275–1281.

3. van Eden, Marc E, Marshall P Byrd, Kyle W Sherrill, and Richard E Lloyd. 2004. “Demonstrating Internal Ribosome Entry Sites in Eukaryotic mRNAs Using Stringent RNA Test Procedures..” RNA 10 (4): 720–730.

4. Bordeleau, Marie-Eve, Ayaka Mori, Monika Oberer, Lisa Lindqvist, Louisa S Chard, Tatsuo Higa, Graham J Belsham, Gerhard Wagner, Junichi Tanaka, and Jerry Pelletier. 2006. “Functional Characterization of IRESes by an Inhibitor of the RNA Helicase eIF4A.” Nature Chemical Biology 2 (4): 213–220.