AS SOON AS scientists began investigating H. pylori, it became clear that strains isolated from different individuals are highly diverse.(A variety of strains can also be found in a single stomach.) Although the strains are identical in appearance, their genetic codes vary greatly. Researchers have determined the complete genomic DNA sequences for two separate H. pylori strains; each has a single small chromosome of approximately 1.7 million nucleotides, comprising about 1,550 individual genes.(In comparison, Escherichia coli—a bacterium inhabiting the intestines—has about five million nucleotides, and humans have about three billion.) Remarkably, about 6 percent of the H. pylori genes are not shared between the two strains, and even the shared genes have a significant amount of variation in their nucleotide sequences. This level of diversity within a species is extraordinary. The genetic differences between humans and chimpanzees—two distinct species—are tiny compared with the differences among H. pyloristrains: 99 percent of the nucleotide sequences in the human and chimp genomes are identical. The substantial variation in H. pylori’s genome suggests that either the bacteria have existed for a very long time as a species or that any particular variant is not so much better adapted to the human stomach as to outcompete all the others. In fact, both statements are true.

My laboratory has identified two particular types of variation. In 1989 we created a library of H. pylori genes by inserting selected fragments of the bacterium’s DNA into cells of E. coli. The E. coli cells can then produce the proteins encoded by the H. pylori genes. We screened the resulting E. coli samples using blood serum from a person (me!) who carried H. pylori in his stomach; because my immune system had been exposed to the bacterium, the antibodies in my serum would