Catecholaminergic Regulation of Hematopoiesis in Mice

To the Editor:

The article by Benestad et al1 published inBLOOD states that murine bone marrow functions are not regulated by neuronal mechanisms. I am surprised by this drastic and final conclusion. Even the title, “No Neuronal Regulation of Murine Bone Marrow Function,” sounds more suitable for a popular press article.

Actually, there are a number of observations that have been disregarded by the investigators and some most recent findings that should be taken in consideration.

(1) A recent publication that has been ignored by the investigators shows that surgical denervation decreases femoral cellularity as well as progenitor cells while mobilizing these cells in the peripheral blood of splenectomized mice. In nonsplenectomized animals, these changes were quickly cleared.2 In addition, Benestad et al1 did find a decreased bone marrow cellularity in chemical sympathectomized mice.

(2) The investigators argue that our results with adrenergic agonists and antagonists that suggest the presence of a catecholaminergic regulation of haemopoiesis are due to nonspecific effects. This seems rather superficial, because, besides the findings quoted, we demonstrated also that (a) bone marrow pre-B cells do express α1B-adrenergic receptors,3-5 (b) that in vitro norepinephrine (NE) and other adrenergic agonists can inhibit myelopoiesis and rescue bone marrow progenitors from the toxic effect of cytotoxic drugs, and that the α1-adrenergic antagonist prazosin neutralizes these effects at concentrations of 10⁻¹⁰ to 10⁻¹² mol/L.3,6 (c) NE protects 77% of mice injected with a supralethal dose of carboplatin (200 mg/kg) and prazosin abolishes the protection. A time-course study showed that this effect was exerted directly on hematopoietic progenitors in the bone marrow.6 

Most recently, we demonstrated that murine bone marrow contains substantial amounts of catecholamines. NE and dopamine (DA) showed a daily rhythmicity, with peak values during the night. The rhythm was disrupted by chemical sympathectomy, whereas epinephrine (E) did not show any rhythmicity or sensitivity to 6-hydroxydopamine. High and low values of NE and DA were associated with high and low values of their metabolites, which indicated a rhythmic catecholamine release. NE but not DA and E was positively associated with the proportion of cells in the G2/M and S phases of the cell cycle. Moreover, NE and DA were found in both short-term and long-term bone marrow cultures as well as in human or murine B-lymphoid cell lines. These findings indicate that endogenous catecholamines in the bone marrow have both neural and cellular origin.5 

In conclusion, the negative results obtained by Benestad et al1 might well reflect a long-term adaptative response of bone marrow cells catecholamines, but in no case can rule out a neuronal influence on hematopoiesis. Finally, as the investigators also beautifully show, bone marrow is richly innervated by afferent adrenergic fibers. Should we argue that such rich innervation is useless?