To the Editor: After not finding a susceptibility gene for schizophrenia, Dr. DeLisi et al. suggested that schizophrenia could result from variation in gene expression. They also recommended that investigators who choose candidate genes for schizophrenia take under consideration various defects observed in schizophrenia. I wish to add three additional recommendations.

First, successful identification of candidate genes for schizophrenia may be enhanced if the higher urban prevalence of schizophrenia is considered. Second, a variation of genetic expression could result from a pathological mechanism known as translational pathophysiology (1). Abnormalities of protein translation could allow genetic, infectious, and nutritional pathways to affect brain development. Third, investigators might consider that a schizophrenia gene enhances survival during prenatal growth.

Translational pathophysiology could result in both enhanced survival and schizophrenia if a gene that provides resistance to an infectious agent through translational interference is expressed during fetal growth and inhibits both viral and host proteins. I hypothesized this Darwinian approach in merging the literature on the flavivirus resistance gene with the geographies of schizophrenia and flaviviruses (2). In this context, the genetic resistance of certain mice to typhoid fever may also have application to schizophrenia. A critical genetic control of typhoid resistance is the natural resistance-associated macrophage protein 1 (NRAMP 1) (3). This gene enhances expression of several proteins, is expressed in neurons, and is associated with behavioral and immune responses to stress (4, 5). The human homologue maps to chromosome 2q35 (3), a location also linked to susceptibility to rheumatoid arthritis (4). Given the controversial hypothesis that persons with schizophrenia are resistant to rheumatoid arthritis, it is noteworthy that chromosome 2 was one of the few chromosomes in which Dr. DeLisi et al. found positive linkages.