A backward view from 16S rRNA to archaea to the universal tree of life to progenotes: reminiscences of Carl Woese.

I first became aware of Carl Woese in the mid-1970s when he and George Fox criticized a few of the 16S rRNA oligonucleotide sequences emerging from Strasbourg in the 10-12 y RNA sequencing project of the first 16S rRNA from Escherichia coli, some of which we were using for assembling RNA binding sites of ribosomal proteins. When I realized that they were attempting to sequence 16S rRNAs from a range of bacteria to classify them phylogenetically, I seriously questioned their sanity. Not because of the goal, which was admirable, but because of the sheer technical difficulty, and slowness, of sequencing large RNA molecules using the original Sanger RNA sequencing method, not to mention the health hazards of regularly preparing rRNA using 20-30 mCi [ (32)P]. My view changed radically, however, with their subsequent prediction of 5S rRNA secondary structures using a phylogenetic approach. Previously, the molecular biology community had been competing to generate the maximum numbers of base pairs in the model RNA molecule E. coli 5S RNA when Fox and Woese introduced the concept of compensatory base changes based on phylogeny for defining secondary structure and applied it to 5S RNA, they found evidence for only about 50% base pairing. This approach had previously been used for tRNA secondary structure predictions but its more general significance had never been acknowledged. Carl subsequently persuaded Harry Noller to apply the same method to predicting secondary structures of the large rRNAs.