In the developing nervous system, precisely patterned connections resu
lt from mechanisms that remodel initially diffuse connections. For exa
mple, ocular dominance column formation depends upon activity-based co
mpetitive interactions. In the developing trigeminal(V) somatosensory
system, injury to afferent inputs prevents somatotopic pattern formati
on; however, afferent impulse blockade does not. What establishes cent
ral V patterns remains unclear. As a first step in assessing the role
of neurotrophins in naturally occurring death of V ganglion cells and
whisker-related pattern formation, the consequences of prenatal NGF in
jections were evaluated. Fetal rats given NGF on both embryonic day (E
) 15 and E18 had 36% more V ganglion cells than normal and lacked whis
ker-related patterns in the V brainstem complex at birth and through p
ostnatal day 3, as determined by cytochrome oxidase histochemistry. Ra
ts injected with NGF on E16 or on E18, or with vehicle had normal gang
lion cell numbers and brainstem patterns. Animals injected with antibo
dies to NGF or an NGF receptor had reduced ganglion cell numbers and n
ormal brainstem patterns. These findings suggest that naturally occurr
ing cell death in the V ganglion is neurotrophically regulated and tha
t this process impacts upon somatotopic pattern formation in the V bra
instem complex. Results of anterograde tracing experiments in NGF-augm
ented animals suggest that pattern disruptions are due to an absence o
f whisker-related patterning in the central projections of V ganglion
cells. Moreover, single primary afferent collaterals labeled by Neurob
iotin injections in the V ganglion did not have widespread or unusuall
y complex arbors. Thus, NGF may affect V pattern formation by preservi
ng or inducing projections to brainstem regions that normally come to
lack such projections, such as the spaces normally demarcating neighbo
ring whisker primary afferent projections.