Vaccines are effective at the onset of a potential infection, but once HIV is established in the body, it seems to be there to stay. The reason for its persistence is latently infected cells — cells that host a copy of HIV’s genetic material, but that are not actively producing new viruses. The improvement of antiretroviral drugs in the 1990s transformed HIV from a death sentence into a manageable disease, but our current medications can only block HIV replication, not root it out from latently infected cells. Satish is also interested in these latently infected cells and what role they play in the course of an HIV infection.

To understand this research, it will help to consider an HIV infection from a tree-thinking perspective:

1. Initial infection. A sexually transmitted HIV infection begins with one virus particle or a few closely related particles. This initial infection is the root of the tree.
2. Initial replication. Within days the virus has begun to replicate exponentially. During this time it is slowly diversifying, evolving through natural selection and genetic drift. The host may feel flu-like symptoms at this point.
3. Invasion of other organs. Within days or weeks, virus particles begin to spread from the blood to other tissues in the body (e.g., the central nervous system, as shown here.). Some of these tissues may be relatively isolated, fostering the formation of genetically distinct clades. Many of these tissues will become reservoirs of latently infected cells.
4. Immune response. When the host’s immune system kicks in, it prevents many viral lineages from replicating, and the flu-like symptoms recede. From the virus’s perspective, this stage looks a bit like a mass extinction! During this purge, natural selection favors variants of the virus that can escape the immune system’s defenses.
5. Drug therapy. Eventually, the virus will bounce back and begin replicating uncontrollably again — unless the person starts antiretroviral therapy. Taking a cocktail of three to five drugs will keep the virus at undetectable levels for an indefinite period of time (though the virus will rebound if the patient stops taking the drugs). Because HIV levels are so low during drug therapy, this period of the infection has always been difficult to study.

This last step was a mystery that interested Satish and his colleagues in San Francisco and Switzerland²: what is going on with HIV while a patient is on antiretrovirals?