Neural tissue from human fetuses is currently used for intracerebral t
ransplantation to treat patients with Parkinson's disease. The develop
ment of the human fetal tissue following grafting has been considered
mostly, up to now, from the neuronal point of view in xenografts. Very
little is known, in contrast, about non-neuronal, glial, or vascular
cells in the grafts. Comparison of the data gathered on the developmen
t of grafted human neurons with those obtained in comparable studies u
sing rat transplants has demonstrated species-specific features. We ha
ve therefore undertaken a series of studies dealing with nonneuronal c
ells in human-to-rat transplants to reveal other possible species-spec
ificity of the human tissue. This study has, ac; cordingly, been devot
ed to the immunohistochemical analysis of microglia of host and donor
origins in a human to rat xenograft paradigm allowing clear distinctio
n of the origin of the cells. Human neural tissue was transplanted as
a cell suspension into the thalamus of adult rats. Amoeboid human micr
oglia were observed in 1-, 2-, and 3-month-old transplants, but their
density, already relatively low at the first stage, decreased further
over time. Ramified human microglia were only occasional. In sharp con
trast, host rat microglia rapidly invaded the transplant in the absenc
e of any sign of necrosis. The rat cells exhibited first an amoeboid m
orphology but progressed at the later stages toward a more mature, ram
ified morphology. These results indicate that donor microglia are quit
e few in number at first and, at least, do not proliferate actively af
ter transplantation. They seem rather to disappear over time. The para
llel migration of a large number of host microglia into the transplant
and the apparent maturation of these cells lead to the formation of a
cellular chimaera. Extrapolation of these results to clinical neural
grafting suggests that most, a not all, microglia in fetal transplants
may rapidly be of host origin in patients.