The concept that bone marrow (BM)-derived cells participate in neural regeneration remains highly controversial and the identity of the specific cell type(s) involved remains unknown. We recently reported that murine BM contains a highly mobile population of CXCR4+ cells that express mRNA for various markers of early tissue-committed stem cells (TCSC), including neural TCSC (

Leukemia 2004:18;29–40
). Here we show that these cells could be isolated from murine BM as a population of Sca-1+lin CD45 cells that not only express neural lineage markers (β-III-tubulin, Nestin, NeuN and GFAP), but more importantly form neurospheres in vitro and cells from these neurospheres after replating into secondary cultures differentiate into neuronal (β-III tubulin+, nestin+) and macroglia (O4+, MBP+, GFAP+) lineages. Our data demonstrate that neural TCSC i) are very rare (0.01% of BMMNC) (ii) present in BM harvested from young (1–2 month) while being barely detectable in older (1-year) mice; iii) are mobilized from BM into peripheral blood (PB) during pharmacological mobilization and iv) are chemoattracted for potential brain regeneration in an SDF-1-CXCR4, dependent manner. In humans the corresponding population of cells is present among CXCR4+CD34+AC133+CD45 BMMNC. Our studies performed on 14 patients with stroke (age 48–75 years) demonstrated in the peripheral blood an increase in the number of CXCR4+CD34+AC133+ cells expressing neural TCSC. The maximal elevation of these cells was observed 24–72 hours after stroke and remained elevated up to one week. The degree of mobilization correlated with younger age, smaller size of the stroke, was best in the patients suffering from a less extensive stroke and worse in patients with dyslipidemias. Thus, we conclude that bone marrow is a potential source of CD45- neural TCSC for brain repair and since purified CD45+ HSC neither express neuronal markers nor differentiate in vitro into neurospheres, we provide for the first time evidence that neural TCSC residing in bone marrow but not trans-dedifferentiated HSC account for neural differentiation of BM-derived cells. Furthermore, our observation that the number of BM-derived neural TCSC decreases with age provides a novel insight into aging and may explain why the brain regeneration process becomes less effective in older individuals. These observations provide rationale for further studies aimed at optimizing therapeutic brain regeneration by BM-derived neural TCSC.

Author notes

Corresponding author

Sign in via your Institution