|Home | About | Journals | Submit | Contact Us | Français|
We thank Küry et al.  for their letter. They raise a number of issues with regard to our initial article that reported on variations in BAK1 gene (MIM# 600516) sequence expressed in the abdominal aorta (AA) tissue when compared with leukocyte genomic DNA sequence [Gottlieb et al., 2009]. In a subsequent letter [Gottlieb et al., 2010], we noted that when we compared the genomic BAK1 sequence in matching blood and tissue samples, the blood and tissue BAK1 genomic sequences were identical to the reference BAK1 genomic sequence, which primarily suggested that an unusual form of RNA editing could possible be involved.
Küry et al.  have raised three issues with regard to possible explanations for our results.
It is clear that the BAK1 sequence we amplified from the cDNA prepared from our DNAase treated RNA (not, of course, RNAase treated, as mistakenly noted by Küry et al. ) resembles both BAK1 and BAK2 yet copies neither exactly. At amino acid 2 the chromosome 20 BAK2 reference sequence of NC_000020.10 has the sequence ATG/GCC/TCG. In contrast, BAK1 mRNA (NM_001188), and all our samples isolated from AAA and all AA cDNA samples have the sequence ATG/GCT/TCG (Table (Table1).1). In Figure 2, Küry et al.  reference five other BAK2 sequences to indicate that there is heterogeneity at nucleotide 217. The results, however, are likely from only three different isolates not five. BAK2 chromosome 20 sequences NC_000020.10 and NT_011362 contain ATG/GCC/TCG, with both sequences having the same citation [Deloukas et al., 2001], suggesting that they are from the same isolate. BAK2 chromosome 20 sequence NG_000850.5 has ATG/GCT/TCG and is referenced in Kiefer et al. . This publication was the original description of the cloning and sequencing of BAK1, BAK2, and BAK3. Here these authors categorize BAK3 as a pseudogene, but indicate that BAK2 is not a processed gene. The chromosome 20 BAK2 sequences of AC_000152.1 and NW_00183664 contain ATG/GCT/TCG from the DNA of Dr. J. Craig Venter [Levy et al., 2007] and so are also from a single isolate. It is interesting to note that Venter's DNA sequence appears to be somewhat of an outlier when compared to the other complete human genomes so far sequenced [Ahn et al., 2009]. Please note that we compared our results to the reference and not ancillary sequences, and that regardless of the sequence variations among the BAK2 genes it is clear that the sequence we amplified contains the sequence present in BAK1.
In Figure 2, Küry et al.  also compared the sequencing results for the same BAK2 sequences and BAK1 sequence at nucleotide 645. All the BAK2 sequences have the CCA/GCG/TGG sequence, whereas the BAK1 sequence is CCA/GCA/TGG. All of our samples are the same as the BAK1 sequence and contain the sequence CCA/GCA/TGG (Table (Table11).
We believe that our critics have been very subjective in their use of sequence data in that they chose not to discuss the concordance of our results at nucleotides 217 and 645 with the BAK1 sequence, but only to discuss the discordance at 217 in the BAK2 sequences by relying on three variant BAK2 sequences. This would appear to be a good example of a selective use of sequence data, so that even though we reported on actual sequence data from individuals, they chose to solely rely on reported data from these three cases.
Although we feel that we are clearly dealing with single nucleotide polymorphisms (SNPs) in BAK1, as noted previously in the original article reporting BAK2 in chromosome 20 [Keiffer et al., 1995], the authors reported that it could not be considered a processed gene. Finally, a recent study has identified correlation of 34 trait-associated SNPs with copy number variations, and interestingly, one of the genes identified was BAK1 [Conrad et al. 2010]. Although the associated trait was testicular germ cell tumors, it suggests that BAK1 SNPs have the potential to be disease associated.
Again, we are grateful to Küry et al.  for raising a number of issues that are likely to be of importance as we uncover increasing amounts of human genome variation. In particular, their letter raises a number of key questions that are only likely to increase in significance in the coming years. These questions include the following:
Finally, The question of prematurely overrating results is clearly a fact of life in modern science. As Küry et al.  noted, it was not us that did so in our article, but rather the popular Scientific Press. The fact that such behavior goes on constantly is perhaps a measure of the importance of public relations (PR) in modern scientific research. All large universities and research institutions now have a PR department, whose goal is to make certain that their research reputation is suitably promoted and enhanced. This has lead to a veritable orgy of press releases announcing countless major scientific breakthroughs, most of which in the harsh light of reality tend to be highly overrated. Further, such behavior is not just limited to the PR departments of institutions, but also the PR departments of some of our most highly rated scientific publications. Perhaps it is time to seriously consider the value of such press releases, as they are in many ways resulting in irresponsible and even poor science. This can perhaps best be illustrated by the numerous press conferences that have announced the results of experiments that have not yet been completed or even undertaken! As scientist and researchers we are constantly being reminded of the positive benefits of publicity, but perhaps it is also time to consider the negative aspects as well.