New findings overturn a major model of where immune memory is stored. Rather than circulating throughout body, as researchers had thought, memory T-cells actually reside in a comfortable niche in the bone marrow waiting for the next chance to fight infection, according to a new article publishing online in Immunity today (May 7th).

"It's very exciting data," said Antonio Lanzaveccia from the Institute for Research in Biomedicine in Switzerland, who was not involved in the research. It would be important to see "what the relationship is between these memory cells and other memory cells that have been described," he added.

Andreas Radbruch from the Charité German Rheumatology Center in Berlin and colleagues wanted to take a closer look at T memory cells because in earlier studies they had found that memory B-cells, which produce antibodies, reside in the bone marrow.

Immunologists have long believed that memory cells come from activated effector T cells that have resigned their ability to fight, and simply remain in circulation until they are re-activated a second time by the same pathogen they initially attacked. Researchers have been able to easily isolate circulating memory T-cells in human subjects, said Lanzaveccia, but accessing bone marrow is much more difficult.

To test whether memory T cells stayed in circulation or retreated to the bone marrow, Radbruch and first author Koji Tokoyoda, also at the German Rheumatology Center, infected a mouse with a pathogen, and then searched for the T-cells specific to that pathogen at different time points. As expected, the specific T-effector cells at first proliferated to fight the infection. Four days in, researcher detected the memory cells mainly located in the lymphoid organs and spleen. But 3-8 weeks after infection -- by which time the remaining cells would have turned into memory cells -- the researchers searched again for those pathogen-specific cells and found that more than 80% of them were now in the bone marrow. These cells remained there for up to 134 days -- the length of time the researchers tested.

To make sure the cells they detected in the bone marrow were "true" memory cells, Radbruch and colleagues characterized their surface molecules and tested them for the characteristics of memory cells: lack of proliferation, decreased gene expression, and the ability to reactivate upon re-infection with the same pathogen they had initially encountered. The cells in the bone marrow passed all three tests. The remaining 20% of cells not in the bone marrow probably belonged to a subset of T-memory cells that had been reactivated or were reacting to chronic infection, Radbruch suggested.

Radbruch and colleagues wanted to find out what attracted the memory cells to the bone marrow. The prime suspect: an adhesion molecule called alpha2 integrin. The molecule showed increased expression on memory cells, and its ligand is predominantly expressed in bone marrow tissue. Also, when the researchers blocked alpha2 integrin with an antibody, the cells no longer homed to the bone marrow.

Researchers went on to characterize the interaction between the memory cells and their bone marrow niche. "What I find totally exciting about this [paper]," said Lanzaveccia, is that "it has very nice and quantitative data." The researchers counted the number of memory cells in the bone marrow, nestled in niches producing nurturing cytokines, to be as high as 5 million. The authors hypothesize that the bone marrow offers memory cells an environment rich in cytokines essential for their survival. IL-7, for example, richly present in bone marrow, helps T-cells that aren't actively proliferating, survive.

The paper poses such a basic question in immunology, Radbruch said, that one of his reviewers had asked why no one has thought to do such an experiment before. "All the good papers are like that," said Lanzaveccia: They make an experimenter think, "Gosh I should have thought of that!"

http://www.cell.com/immunity/abstract/S1074-7613(09)00187-3