G-CSF protects motoneurons against axotomy-induced apoptotic death in neonatal mice

Abstract

Background: Granulocyte colony stimulating factor (G-CSF) is a growth factor essential for generation of neutrophilic granulocytes. Apart from this hematopoietic function, we have recently uncovered potent neuroprotective and regenerative properties of G-CSF in the central nervous system (CNS). The G-CSF receptor and G-CSF itself are expressed in alpha motoneurons, G-CSF protects motoneurons, and improves outcome in the SOD1(G93A) transgenic mouse model for amyotrophic lateral sclerosis (ALS). In vitro, G-CSF acts anti-apoptotically on motoneuronal cells. Due to the pleiotrophic effects of G-CSF and the complexity of the SOD1 transgenic ALS models it was however not possible to clearly distinguish between directly mediated anti-apoptotic and indirectly protective effects on motoneurons. Here we studied whether G-CSF is able to protect motoneurons from purely apoptotic cell death induced by a monocausal paradigm, neonatal sciatic nerve axotomy.

Results: We performed sciatic nerve axotomy in neonatal mice overexpressing G-CSF in the CNS and found that G-CSF transgenic mice displayed significantly higher numbers of surviving lumbar motoneurons 4 days following axotomy than their littermate controls. Also, surviving motoneurons in G-CSF overexpressing animals were larger, suggesting additional trophic effects of this growth factor.

Conclusions: In this model of pure apoptotic cell death the protective effects of G-CSF indicate direct actions of G-CSF on motoneurons in vivo. This shows that G-CSF exerts potent anti-apoptotic activities towards motoneurons in vivo and suggests that the protection offered by G-CSF in ALS mouse models is due to its direct neuroprotective activity.

Figure 1

The G-CSF receptor is expressed by motoneurons in the spinal cord of wt neonatal mice. (A-C) Double fluorescence immunostaining of G-CSF receptor and CHAT (choline acetyltransferase) in the neonatal mouse spinal cord ventral horn. CHAT is used as a marker for motoneurons. All defined α motoneurons express G-CSFR (indicated by an arrow). All photomicrographs with 40× original magnification (OM), size bar 25 μm.

Figure 2

G-CSF receptor expression after sciatic nerve axotomy. (A) G-CSF receptor is upregulated after sciatic nerve axotomy in spinal cord of neonatal mice. Quantitative PCR for G-CSF receptor of spinal cords of axotomized or not neonatal mice (induction of 55%; n = 3, **p < 0.005). (B) G-CSF receptor staining intensity is stronger on the motoneurons after axotomy. Fluorescent intensity of G-CSF receptor staining on motoneurons was calculated on the ipsi- and contralateral side with an image processing program (*p < 0.05, n = 49).

Figure 3

G-CSF concentration in the CNS. Levels of G-CSF are higher in the brain and in the spinal cord of mice overexpressing G-CSF (BEG -/T CaMK -/T) compared to wild type littermate controls (BEG -/- CaMK -/-). G-CSF proteins were quantified by ELISA with protein lysates from brain or spinal cord (n = 3, *p < 0.05).

Figure 4

Examples of the histological evaluation of CHAT positive cells in the ventral horn of the lumbar spinal cord of neonatal mice after sciatic nerve axotomy. Sections from the lumbar spinal cord were stained with CHAT antibodies. (A, B) Examples of motoneurons in the ventral horn contralateral and ipsilateral to the lesion side in BEG -/- CaMK -/- mice (wild type littermates). (C, D) Examples of lumbar sections contra- and ipsilateral to the lesion from a BEG -/T CaMK -/- littermate (non-G-CSF expressing). (E, F) Examples of lumbar sections contra- and ipsilateral to the lesion from a G-CSF overexpressor (BEG -/T CaMK -/T mouse). (B, D, F) There are more CHAT-positive cells detectable 4 days after axotomy on the ipsilateral side in mice overexpressing G-CSF (BEG -/T CaMK -/T). All photomicrographs with 20× original magnification (OM), size bar 50 μM.

Figure 5

Overexpression of G-CSF improves motoneuron survival following sciatic nerve axotomy in neonatal pups. (A) Quantification of the absolute number of α; motoneurons per section of the lumbar spinal cord on the contralateral (unlesioned) side. Numbers are not significantly different between the three genotypes studied. (B) On the axotomized side numbers of detectable motoneurons are dramatically decreased in the littermate animals lacking the CaMK-tta driver transgene (wt, BEG -/- CaMK-/-; and BEG -/T CaMK-/-), whereas overexpression of G-CSF in the CNS results in significantly higher numbers of surviving motoneurons (***p < 0.0001 by ANOVA and Fisher's LSD post-hoc analysis). (C) Percentage of surviving motoneurons on the ipsilateral (axotomized) relative to the corresponding contralateral (control) side. While only 50 - 60% of the motoneurons survive in the genotypes lacking the CaMK-tta driver transgene, close to 90% of the motoneurons survive in the presence of the driver (i.e. in animals overexpressing G-CSF) (***p < 0.0001 by ANOVA and Fisher's LSD post-hoc analysis). (Number of animals: wt, n = 13; BEG -/T CaMK -/-, n = 6; BEG -/T CaMK -/T, n = 7; 6 sections were studied per animal).

Figure 6

Size distribution of surviving neurons in the spinal cord of neonatal mice following sciatic nerve axotomy. On the contralateral side (hatched bars) the size distribution between the three genotypes is similar. On the ipsilateral side (filled bars) there is a downward shift in the mean size of the surviving motoneurons in the littermates without overexpression of G-CSF (BEG -/- CaMK -/-; BEG -/T CaMK -/-). (p < 0.0001 for BEG -/T CaMK -/T vs. BEG -/- CaMK -/-; **p = 0.0008 for BEG -/T CaMK -/T vs. BEG -/T CaMK -/-; ANOVA followed by Fisher's LSD; Number of motoneurons measured: wt, n = 301; BEG -/T CaMK -/-, n = 197; BEG -/T CaMK -/T, n = 227).