Reader Comments

Post a new comment on this article

EBNA1 Gene Operates at Nucleic Acid, not Protein, Level

Posted by forsdyke on 08 Nov 2010 at 20:59 GMT

Notwithstanding the elegent experiments in the present paper that are held to support the view of the authors (and many others) that MHC presentation of EBNA1 peptides is inhibited by a process operating primarily at the level of the EBNA1 protein, I believe that insufficient attention is given to a view advanced by Cristillo et al. 2001 [1] and supported by Tellem et al. 2008 [2].
Focussing on the translation process required to generate the peptide in the first place, the present authors assert that "the effect is mediated by the peptide and not the RNA sequence." They refer to their earlier paper [3]. Here they criticized the Tellem study [2] where the nucleic sequence had been changed without changing the protein sequence. Tellem found that this removed the inhibition of MHC peptide presentation.
Apcher et al. do not refer to the converse experiment of "Nature" where the protein sequence was changed without changing the nucleic acid sequence. By inserting or subtracting a single base at the right position in a coding sequence one can change a reading frame, thus drastically changing the amino acid sequence of the protein, but only marginally (one base) changing the nucleic acid sequence. That Nature has already performed this experiment was discovered by Zaldumbide et al. [4]. In this case inhibition of MHC peptide presentation remained. The case was as follows:
DNA sequence similarities indicate that EBV (human herpes virus 4) and KSHV (human herpes virus 8) are homologous (evolved from a common ancestor). The function of the EBNA1 protein in EBV is, according to the conventional wisdom, similar to that of the LANA1 protein in KSHV. So they would be expected to have similar protein sequences. But they do not. They are dramatically different.

Nevertheless, the corresponding genes retain homologous nucleic acid sequences in the repeat region. Zaldumbide et al. [4] showed that the difference in protein sequences can be accounted for by a reading frame shift. An interpretation of this is that the nature of the encoded protein is largely irrelevant for the inhibition of the display of peptides for T cell activation. It is the nucleic acid seqence that is preserved because selection has acted at the nucleic acid level, not at the protein level.

Cristillo et al. [1] suggested that the actual synthesis of the EBNA1 (or LANA1) proteins was of less importance than the permissive signalling that would allow peptides from these proteins to associate with MHC in a form that could be recognized by T-lymphocytes. This permissive signally required the formation of double-stranded RNA (dsRNA). The purine-loading would have militated against such signalling by diminishing the probability of formation of dsRNA. Thus, the effect was not due to RNA structure per se (e.g. stem-loop pattern), but due to formation of a sufficient length of dsRNA (by mechanisms for which there is growing evidence [5]).

1. Cristillo AD, Mortimer JR, Barrette IH, Lillicrap TP, Forsdyke, DR (2001) Double-stranded RNA as a not-self alarm signal: to evade, most viruses purine-load their RNAs, but some (HTLV-1, Epstein-Barr) pyrimidine-load. J Theor Biol 208: 475-491.

2. Tellem J, Smith C, Rist M, Webb N, Cooper L, Vuocolo T, Connolly G, Tscharke DC, Devoy MP, Khanna R (2008) Regulation of protein translation through mRNA structure influences MHC class 1 loading and T cell recognition. Proc Natl Acad Sci USA 105: 9319-9324.

3. Apcher S, Komarova A, Daskalogianni C, Yin Y, Malbert-Colos L, Fahraeus R (2009) mRNA translation regulation by the Gly-Ala repeat of Epstein-Barr virus nuclear antigen 1. J Virol 83: 1289-1298.

4. Zaldumbide A, Ossevoort M, Wiertz EJHJ, Hoeben RC (2007) In cis inhibition of antigen processing by the latency-associated nuclear antigen 1 of Kaposi sarcoma Herpes virus. Mol Immunol 44: 1352-1360.

5. Forsdyke DR (2011) Evolutionary Bioinformatics. 2nd Edition. Springer, New York.

No competing interests declared.

RE: EBNA1 Gene Operates at Nucleic Acid, not Protein, Level

fahraeus replied to forsdyke on 09 Nov 2010 at 09:51 GMT

Response to comments by Forsdyke.

How can you say that the "EBNA1 operates on the nucleic acid level, and not the protein level", when the published results suggest the opposite and when you in fact have no results of you own to support this conclusion?

When we first published that the Gly-Ala repeat sequence of the EBNA1 protein suppresses its own mRNA translation in cis back in 2003, you contacted me to say that this publication supported your view since long on the effect of purine rich sequences on mRNA translation. This was despite the fact that we had shown that antibodies against the Gly-Ala could partly overcome its effects in vitro. Hence, I was not entirely sure about you RNA hypothesis, even though interesting, and I was concerned that you did not seem to have taken any notice of this observation. Since then you are vigorously defending your model and you recently contacted me with obvious concerns that our latest MS in PLoS Pathogens might put a dent in your theory. I say this because I find your statement "no conflict of interest" not entirely correct and I do not find your comments very objective either.

In your comments to our PLoS Pathogen paper you omit to mention the important result that changes in the 5' UTR of a Gly-Ala-carrying mRNA is sufficient to overcome its suppressing effect on mRNA translation. This is well described in the MS and in a previous MS by Apcher et al in J. of Virol and, surprisingly, even though I mentioned this in our e-mail correspondence you do not to include this in your comments. Why? In your model the RNA sequence of the encoding Gly-Ala would prevent the ribosome from reading through this sequence but you completely fail to offer an explanation to how changes in the 5' UTR, while leaving the coding sequence intact, would fit with this model. If the 5' UTR sequence had any effects on the Gly-Ala encoding RNA structure, would this not also happen if it was located in the 3' UTR? It does not. Furthermore, as the GAr is located app. 100 amino acids downstream of the first AUG, your model would predict that there would be a smear of truncated EBNA1 products expressed in EBV-infected cells. There are no such products. And how do you explain that the Epstein-Barr virus would rely on a mechanism to suppress the production of antigenic peptides that give rise to truncated products? It seems less constrain to me that the Gly-Ala peptide inhibits translation initiation in cis, as the published results in fact suggest.

You also argue that since the LANA1 protein has been proposed to do the same trick as EBNA1 but as they share little amino acid sequence similarities the peptide sequence cannot carry out the effect. This is the experiment that you refer to that "nature" has provided us with. This logic is flawed. Just because two peptide sequences are not identical does not prove that they cannot mediated similar effects. We never claim that the Gly-Ala peptide is the only peptide sequence that might have these properties. Why would it? It is possible, and likely, that other repeats will have the same effect, perhaps not even via the same mechanisms.

You also refer to a previous MS by Tellam et al that support the view that the RNA sequence suppresses ribosomal read-through. I discussed in our most recent e-mail correspondence why I believe that Tellam and us come to different conclusions but this did apparently not impress on you as you left this out in your comments as well. I will not go into this discussion here and anyone interested can read this in previously published papers.

Like I mentioned to you in our previous correspondence, I do not mean to say that your model is wrong in general but I believe there is a strong physiological and experimental data to support that in the case of the Gly-Ala your model is incorrect. I wish you would be a bit more open minded, look at the results published and ask yourself if there might not just be a slight possibility that your model is not as general and unifying as you wish? This reflects my main concern -you push the generalization of your model so hard that you fail to reflect on, or see, results that offer alternative explanations.

Competing interests declared: I am the corresponding author of the article in question.

RE: RE: EBNA1 Gene Operates at Nucleic Acid, not Protein, Level

forsdyke replied to fahraeus on 10 Nov 2010 at 14:53 GMT

Dr. Fahraeus has misinterpreted the substance and intent of my comment. For a perhaps clearer outline of my suggestion please see the Nov. 2010 End Note at:

No competing interests declared.