The purpose of this study was to develop a novel therapy for Parkinson's disease (PD). We recently reported that dextromethorphan (DM), an active ingredient in a variety of widely used anticough remedies, protected dopaminergic neurons in rat primary mesencephalic neuron-glia cultures against lipopolysaccharide (LPS)-mediated degeneration and provided potent protection for dopaminergic neurons in a MPTP mouse model. The underlying mechanism for the protective effect of DM was attributed to its anti-inflammatory activity through inhibition of microglia activation. In an effort to develop more potent compounds for the treatment of PD, we have screened a series of analogs of DM, and 3-hydroxymorphinan (3-HM) emerged as a promising candidate for this purpose. Our study using primary mesencephalic neuron-glia cultures showed that 3-HM provided more potent neuroprotection against LPS-induced dopaminergic neurotoxicity than its parent compound. The higher potency of 3-HM was attributed to its neurotrophic effect in addition to the anti-inflammatory effect shared by both DM and 3-HM. First, we showed that 3-HM exerted potent neuroprotective and neurotrophic effects on dopaminergic neurons in rat primary mesencephalic neuron-glia cultures treated with LPS. The neurotrophic effect of 3-HM was glia-dependent since 3-HM failed to show any protective effect in the neuron-enriched cultures. We subsequently demonstrated that it was the astroglia, not the microglia, that contributed to the neurotrophic effect of 3-HM. This conclusion was based on the reconstitution studies, in which we added different percentages of microglia (10-20%) or astroglia (40-50%) back to the neuron-enriched cultures and found that 3-HM was neurotrophic after the addition of astroglia, but not microglia. Furthermore, 3-HM-treated astroglia-derived conditioned media exerted a significant neurotrophic effect on dopaminergic neurons. It appeared likely that 3-HM caused the release of neurotrophic factor(s) from astroglia, which in turn was responsible for the neurotrophic effect. Second, the anti-inflammatory mechanism was also important for the neuroprotective activity of 3-HM because the more microglia were added back to the neuron-enriched cultures, the more significant neuroprotective effect was observed. The anti-inflammatory mechanism of 3-HM was attributed to its inhibition of LPS-induced production of an array of pro-inflammatory and neurotoxic factors, including nitric oxide (NO), tumor necrosis factor alpha (TNF-alpha), prostaglandin E2 (PGE2) and reactive oxygen species (ROS). In conclusion, this study showed that 3-HM exerted potent neuroprotection by acting on two different targets: a neurotrophic effect mediated by astroglia and an anti-inflammatory effect mediated by the inhibition of microglial activation. 3-HM thus possesses these two important features necessary for an effective neuroprotective agent. In view of the well-documented very low toxicity of DM and its analogs, this report may provide an important new direction for the development of therapeutic interventions for inflammation-related diseases such as PD.