Granulocyte Colony-Stimulating Factor (G-CSF) Stimulates Sympathetic Nervous System Activity in the Bone Marrow

Abstract

Granulocyte colony-stimulating factor (G-CSF) is the agent most commonly used clinically to elicit hematopoietic stem and progenitor cells (HSPCs) mobilization from the bone marrow (BM) to the peripheral blood. Our previous studies indicate the sympathetic nervous system (SNS) controls G-CSF-induced HSPCs egress from the BM and that G-CSF decreases catecholamine levels in bone/BM tissues (Cell 2006; 124:407). However the molecular basis for this depletion is unknown. Here we show that G-CSF stimulates selectively the activity of the SNS in bone/BM microenvironment in mice. Norepinephrine (NE), a catecholamine and the principal neurotransmitter of the SNS, is synthesized by the rate-limiting enzyme tyrosine hydroxylase (TH). Since TH is a catecholaminergic marker which reflects the extent of sympathetic innervation, we evaluated its expression in bone/BM tissues after G-CSF treatment. TH immunofluorescence staining analysis in longitudinal frozen sections of femurs exhibited a strong punctuate staining of the catecholamine-producing enzyme that was increased 10-fold in G-CSF-treated mice compared to control animals. This result was also confirmed by Western blot analysis of equal amount of protein obtained from BM lysates of G-CSF-treated and control animals. Because NE release and removal from the extracellular space determine tissue NE turnover and regulate the sympathetic activation, we evaluated NE release and uptake in sympathetic cervical ganglia (SCG) in organ cultures treated with G-CSF. Although the release of [3H]NE from cultured ganglia was not affected by G-CSF, we found a significant reduction in the reuptake rate. NE clearance from the synaptic cleft is mediated by the Na+-dependent NE transporter (NET). NET is selectively expressed and localized in lipid rafts of the plasma membrane of the noradrenergic nerve terminals and it is internalized as a result of its down-regulation. We thus isolated lipid rafts from BM homogenized in lysis buffer containing 1% Triton X-100 and subjected to sucrose gradient centrifugation. We found that G-CSF induced the redistribution of NET from the plasma membrane as revealed by a strong reduction of NET levels in BM lipid raft-enriched gradient fractions and by an increase of NET levels in BM non-raft fractions following G-CSF treatment. Moreover Q-PCR analysis of catecholamine-inactivating enzymes revealed reduced mRNA levels of the monoamine oxidase (MAO) and the catechol-O-methyltransferase (COMT) enzymes derived from BM of mice injected with G-CSF compared to untreated animals. Altogether, these results demonstrate that G-CSF treatment affects NE turnover impairing the removal of NE from the synaptic cleft and therefore leading to an increase of the concentration of the extracellular NE. A likely consequence is a prolongation of the postsynaptic action of NE on its receptors and ultimately an augmented sympathetic activity in the bone marrow microenvironment.