• Greek proteomics study on effect of cell phone radiation on mice brain

Few days ago research group from Greece published study that examined effects of cell phone radiation and DECT phone radiation on mice brain using proteomics approach. After exposure to radiation, that was generated by regular phones connected to network and playing music, mice brains were examined for changes in expression levels of hundreds of proteins.

Greek researchers concluded that both cell phone radiation and DECT phone radiation alters amounts of over 100 proteins in mice brain. In the press release researchers pointed out that the affected proteins are important in regulation of learning, memory and in regulation of processes leading to Alzheimer’s disease.

Unfortunately, the Greek study falls far from the target.

The only effect that was somewhat shown is the possibility that cell phone radiation might alter levels of some brain proteins. The indication of the processes that might be affected was a pure speculation because the data obtained from the proteomic analysis were insufficiently confirmed. Certain proteins were named as affected but the confirmation experiments that the change is real were not performed.

Analysis of changes in protein expression was performed using 2D-electrophoresis followed by detection of proteins by coomassie blue staining. Thereafter, the images of stained gels were analyzed using the PDQuest software. This is already “obsolete” approach.

• Execution of proteomic analysis by relying on manual matching of the protein spots in 2D-gels is inaccurate. This approach was justified, though already outdated, few years ago. Currently the method of choice, that secures accuracy of matching, is the 2D-DIGE method where internal standard, exposed sample and sham sample, labeled with different stains, are run in the same gel. This assures that the same protein in exposed and in sham sample will localize to the same spot on the 2D-gel.
• Because of the above, the number of the affected proteins indicated in the study might be very inaccurate. Especially, that some of the protein spots seen in the representative gel are very faint and some areas where the “affected” proteins are located, the background is very high. These increase the inaccuracy of manual matching between exposed and control samples.
• Internal standard in 2D-DIGE, that is the same for all gels makes possible to compare protein amounts between gels what assures reliability of the analyses. Lack of such uniform internal standard in this study makes analysis of the expression changes inaccurate.

Table 2 of the article lists all affected proteins and informs whether there was increase or decline in the expression of protein.

• Unfortunately, the authors did not mention the ratios of increase/decline in protein expression for each of the proteins. This information would be very valuable in estimating potential effects on signaling pathways.

The changes in expression of proteins detected using 2D-electrophoresis were not confirmed with other methods, like western blot (except for a couple of proteins). The information obtained by 2D-electrophoresis is very inaccurate. The changes in the expression of proteins must be confirmed. Only 2D-electrophoresis data are insufficient even if the statistical analysis shows changes as significant. Out of the over 100 proteins detected as responding to the exposure many will be false positives. Especially, because of the limitations of 2D-proteomic approach described above (spot size, background, manual matching).

Because it is not known which of the proteins listed in Table 2 are really affected, the pathway network generated with their use by Ingenuity Pathway Analysis is useless. Many of the proteins included in it might be false findings and, consequently, the pathways finds – false. And we do not know which are which.

Because of the inaccuracy of 2D-proteomics used in this study many of the “affected” proteins might be false observations whet leads, in turn, to detection of false pathways using Ingenuity Pathway Analysis. This causes that the whole discussion about the potential impact of the “observed changes” in certain protein expression on the brain physiology and on the potential health risks is baseless. It is not supported by the experimental data presented in the study.

One can wonder why reviewers of this study did not request additional experiments or shortening of the manuscript by removing all speculative, not supported by the data, content of the discussion. The study in its current format should be not published. At least confirmatory experiments showing that the 2D-detected changes are real are must, in this kind of study.

It is certainly study that goes in the right direction, examining molecular level changes in animal brain exposed to cell phone radiation. Unfortunately, the execution of this study prevents drawing any conclusions concerning the possible health effects of this radiation.

The risks listed in the press release are not supported by the experimental evidence as presented in the study.

Because of the “grave consequences” of the exposure, presented in this study but not supported by the data, it would be prudent if the authors would either immediately publish erratum or withdraw the study and perform additional experiments.