However, almost as soon as this "discovery" was announced, the scientific community began to express doubts as to the nature of the NASA group's experiment and their interpretation of the results. Public discussion of this soon faded into the background, but it remained an active debate in the scientific community as the publication of two recent papers in Science demonstrates. Both papers refute the NASA claim that the microbe was incorporating arsenic in its DNA and both suggest that the NASA study was compromised by potential contamination. In other words, the results from a sloppy experiment were rushed into print, press release, and press conference before anyone bothered to double check the results.
The lead author of the NASA study, Felisa Wolfe-Simon, responded to the new papers saying, "there is nothing in the data of these new papers that contradicts our published data." Apparently, English is Dr. Wolfe-Simon's second language because two papers claiming that your results were due to contamination is pretty much contradicting your data.
Abstracts of the two Science papers can be found here and here (Since Science is a "real", peer-reviewed scientific journal, you have to pay to read the full paper).
Why make such a big deal over sketchy research that hasn't been confirmed? Why go public with results without getting comments from your scientific peer group? Quite simple: Someone's funding was coming up for review and what better way to demonstrate how important your research is than to hold a press conference to announce a "breakthrough". In truth, this sort of thing happens all the time; probably 90% of the scientific articles you read in the popular press are a result of these fluffy, attention-whore press releases. It goes without saying that the bureaucrats at NASA that make the funding decisions lack the technical skills to critically evaluate Wolfe-Simon's work, so to them, a press release is just as good a sign of "progress" as a paper in a refereed journal, besides a press release is shorter and not filled up with all that technical stuff.
Sic transit gloria scientia...