When molecular or cell biologists express their scorn for classic biology, it is near-inevitably anatomy that they bring up: "Counting bug legs" is favorite shorthand for the modern biologist to express how dull and irrelevant classic biology is to a true understanding of biological processes.

But molecular biology now has been around long enough to get some of the same treatment.

"A lot of people have the concept of what I call '96-well plate immunology,'" Jonathan Bromberg told BioWorld Today. Roughly speaking, he is referring to putting different mixtures of cell types and molecular factors across the 96 wells, screening the interactions, and presto! - a scientific paper. Or a drug target.

Bromberg, who is a professor at the Mount Sinai School of Medicine in New York, said that to make any headway in his field of study - how the immune system reacts to organ transplants and ultimately, how to prevent organ rejection - "we have to think in terms of anatomy and cell movement, rather then 96-well-plate immunology."

And he's got the data to back him up. In the June 2006 issue of Nature Immunology, senior author Bromberg and his colleagues at Mount Sinai, the National Institute of Allergy and Infectious Diseases in Bethesda, Md., and Dresden University of Technology in Germany show that in mice, one big difference between organ transplant rejection and tolerance is the different pathway taken by the immune system cells that present the donor antigens: into the lymph nodes for tolerance, into the spleen for rejection.

The scientists transplanted a donor heart into mice and investigated the immune system's response to the organ, both in cases where it was tolerated and in cases where it was rejected. They used a monoclonal antibody to identify the cell population that acquires antigens from the donor hearts and presents them to the recipient immune system for either tolerance or rejection. They found that a subset of dendritic cells known as plasmacytoid dendritic cells presented the antigens in both cases.

The finding that plasmacytoid dendritic cells present the antigens was the first surprise in a study that, Bromberg said, was full of them. But the plasmacytoid dendritic cells showed very different behavior depending on whether the transplant was tolerated or rejected. For tolerance, the plasmacytoid dendritic cells migrated mainly back into lymph nodes, where they interacted with T cells to induce the development of regulatory T cells, which suppress the immune system.

Mice rejecting the transplants had hardly any dendritic cells in the lymph nodes. Instead, in those animals, the antigen-presenting cells migrated preferentially into the spleen, where they could not interact with T cells to generate the regulatory T cells that appear to be responsible for graft tolerization.

Transplanting dendritic cells from the lymph nodes of tolerant animals to mice with a transplant from a different heart provided the recipients of the dendritic cells to second-order tolerance of the transplanted organ. While the graft was not accepted permanently, it was tolerated significantly longer than in control mice. While the idea that tolerance to one graft can prolong the survival of a different graft in a different animal - after all, the whole problem underlying transplant rejection in the first place is that donors, and hence their organs, each have unique antigens - Bromberg explained that "since we were using large numbers of genetically inbred strains of mice, the heart that the second mouse received was genetically and therefore antigenically identical to the heart that the first mouse received."

Unfortunately, in human transplant recipients, the lymph pathway is the road less traveled by the dendritic cells. Though Bromberg said that "dozens" of tolerogenic protocols exist in animals, none of them work well enough in humans to have a clinical impact on transplant rejection. For now, the clinical approach is immunosuppression, add some immunosuppression and then perhaps more immunosuppression.

That approach, Bromberg said, is partly a consequence of the 96-well-plate approach. "We have potentially a thousand drug targets," he said. "But it's not feasible to give a patient a thousand drugs." Clinically, another approach is more likely to lead to ultimate success.

"Let's look at our existing molecules, but ask some different questions," he said.