Stephen C. Harrison

Stephen C. Harrison

Giovanni Armenise - Harvard Professor of Basic Biomedical Science
Stephen C. Harrison


Influenza virus in humans evolves in response to pressure from immunity in the potentially susceptible population, resulting in progressive variation of viral antigenicity ("antigenic drift").  Introduction into human circulation of a new serotype from avian or swine hosts ("antigenic shift") escapes immunity in most individuals, usually leading to a pandemic that re-initiates the process of antigenic drift from a fresh starting point.  Reactivation of B-cell memory appears to dominate, over activation of naive B cells, in the response to drifted variants.  Subsequent affinity maturation is thus part of an "arms race" between evolution of the circulating virus in response to population immunity and evolution in an individual of antibody lineages in response to infection by drifted variants.  A consequence of the dominance of memory recall upon exposure to drifted variant is that an individual's initial exposure influences any later response to a closely related viral antigen ("immune imprinting").  Our current efforts, in collaboration with groups at Duke University, Ragon Institute, and Dartmouth College, seek answers to the following questions.  (1) How does initial exposure by vaccination (four protein components derived from an H1 subtype of influenza A, an H3 subtype of influenza A, and two lineages of influenza B) compare with initial exposure by infection (a single subtype)? (2) When an individual imprinted on one subtype encounters the antigen (as vaccine or as virus) of another subtype, what is the balance, in the resulting memory B cell compartment, between reactivated, cross-reactive B cells (potentially further affinity matured) and affinity-matured descendants of newly elicited, naive B cells. (3) What are the structural correlates of the answers to questions (1) and (2)?  How different (both in local structure and in affinity for pre-existing antibodies) does an epitope need to be before it is seen as "new" by an immune system imprinted on a previously circulating subtype (e.g., in the 1968 transition from H2 to H3 or in the encounter of an H1 by vaccination of someone imprinted by early infection with H3).  These questions are all formulated in terms of the principal surface antigen of influenza virus (its hemagglutinin, HA).  We will also investigate parallel questions in the context of the second surface antigen on the virus (its neuraminidase, NA).

For more information about our work on influenza immunity and its relevance for vaccine design, see For our research on virus entry and virus structure, see


Contact Information

Harvard Medical School
Seeley G Mudd Building, Room 130
250 Longwood Avenue
Boston, MA 02115
p: 617-432-5607

Community or Program Affiliation