Detecting a Bacterial Protein to Understand Cancer Risk

Abstract

Featured Article: Blaser MJ, Perez-Perez GI, Kleanthous H, Cover TL, Peek RM, Chyou PH, et al. Infection with Helicobacter pylori strains possessing cagA is associated with an increased risk of developing adenocarcinoma of the stomach.

In 1982, Helicobacter pylori was discovered in the human stomach in association with inflammatory cells infiltrating the gastric mucosa, a condition known as chronic gastritis. Barry Marshall and Robin Warren re- ceived the 2005 Nobel Prize in medicine for the discov- ery of H. pylori and determining its role in peptic ulcer disease. Within several years, it became clear that per- sons carrying H. pylori were also at increased risk for the most prevalent types of gastric cancer (1). Yet, H. pylori colonization is highly prevalent, and only a frac- tion of colonized individuals become ill. Work in my laboratory since 1985 has sought to define the antigens of H. pylori, to identify virulence factors, and to de- velop diagnostic tests to ascertain carriage and geno- type. Studies initiated in 1988 by postdoctoral fellow Dr. Timothy Cover focused on an activity in culture supernatants from some but not all H. pylori strains that induced vacuole formation in epithelial cells. Cover ultimately purified a protein, which we called VacA, that specifically signals host cells.

In the 1980s, to characterize the vacuolating activity Cover prepared culture supernatants from toxin-positive and toxin-negative strains, which we probed in Western blot analyses with serum from persons carrying such strains. The strategy worked! We identified a band mi- grating at 128 kDa that was recognized by serum IgG from persons with toxin-positive strains but not from those with toxin-negative strains (2). Importantly, only approximately 60% of persons with gastritis showed anti- bodies to this 128-kDa protein, in contrast to 100% of patients with duodenal ulceration. The following year, when Crabtree et al. reported that gastric IgA antibodies recognized a 120-kDa band in nearly identical proportions of patients with gastritis and ulcer disease (3), we knew that we were on the right track!

We also screened libraries of H. pylori genes in bacteriophage )\ gt-11 by using serum from a person (me) who had strong antibody responses to H. pylori. This approach yielded Escherichia coli clones that pro- duced H. pylori antigens recognized by my serum IgG. In July 1989, we purified a clone with a high molecular weight antigen. Dr. Murali Tummuru, another post- doctoral fellow in the lab, completed the cloning and found a 128-kDa protein product (4), which we dis- covered to be the same protein that Tim Cover studied. We initially called the gene tagA (toxin-associated gene A); however, we learned that Italian colleagues (Anto- nello Covacci, Rino Rappuoli, and others) had discov- ered the same protein (5), which they had intended to call caiA (cytotoxin-associated immunodominant gene A). We compromised and coined a single gene name (cagA) that has persisted into the present (4, 5).

Our work and that of Crabtree et al. (2, 3) provided evidence that carrying cagA-positive strains increased the risk of peptic ulcer disease. We then focused on gastric cancer to determine whether cagA positivity was associ- ated with an increased risk for that disease. We used a recombinant fragment (orv660) as an antigen to detect specific serum anti-CagA IgG (6). We validated the assay with serum from persons who were known to be H. pylori negative or who were H. pylori positive but from whom a cagA-negative strain was isolated. These studies showed a strong specificity for the assay, and nearly all those who had a cagA-positive strain were seropositive (indicating the high sensitivity of the assay).

With Dr. Abraham Nomura, we had previously found an association between H. pylori infection and intestinal-type distal gastric cancer in Japanese Ameri- can men in Hawaii (1). Using the now-validated re- combinant CagA assay, we assessed CagA associations with such cancers. Our studies showed that H. pylori– positive men who carried a cagA-positive strain had a risk of developing intestinal-type distal gastric adeno- carcinoma in the subsequent 21 years that was in- creased by 130% (odds ratio, 2.3; 95% CI, 1.0 –5.2), compared with men carrying a cagA-negative strain (6). Thus, the polymorphisms of H. pylori permitted the development of an assay that detects serum responses to a specific bacterial protein and is thereby able to predict the risk for developing the most com- mon form of gastric cancer worldwide.

Much subsequent work has confirmed our obser- vations, and we now know that H. pylori injects the CagA protein into its host’s epithelial cells! The in- jected CagA interacts with host proteins to determine cell properties and fate. Our work that used host anti- body responses detected in clinical samples to probe bacterial antigens has advanced our understanding of gastric carcinogenesis and ultimately opened the door to exploring whether other important clinical condi- tions are influenced by CagA status.