Granulocyte colony stimulating factor attenuates inflammation in a mouse model of amyotrophic lateral sclerosis

Abstract

Background: Granulocyte colony stimulating factor (GCSF) is protective in animal models of various neurodegenerative diseases. We investigated whether pegfilgrastim, GCSF with sustained action, is protective in a mouse model of amyotrophic lateral sclerosis (ALS). ALS is a fatal neurodegenerative disease with manifestations of upper and lower motoneuron death and muscle atrophy accompanied by inflammation in the CNS and periphery.

Methods: Human mutant G93A superoxide dismutase (SOD1) ALS mice were treated with pegfilgrastim starting at the presymptomatic stage and continued until the end stage. After long-term pegfilgrastim treatment, the inflammation status was defined in the spinal cord and peripheral tissues including hematopoietic organs and muscle. The effect of GCSF on spinal cord neuron survival and microglia, bone marrow and spleen monocyte activation was assessed in vitro.

Results: Long-term pegfilgrastim treatment prolonged mutant SOD1 mice survival and attenuated both astro- and microgliosis in the spinal cord. Pegfilgrastim in SOD1 mice modulated the inflammatory cell populations in the bone marrow and spleen and reduced the production of pro-inflammatory cytokine in monocytes and microglia. The mobilization of hematopoietic stem cells into the circulation was restored back to basal level after long-term pegfilgrastim treatment in SOD1 mice while the storage of Ly6C expressing monocytes in the bone marrow and spleen remained elevated. After pegfilgrastim treatment, an increased proportion of these cells in the degenerative muscle was detected at the end stage of ALS.

Conclusions: GCSF attenuated inflammation in the CNS and the periphery in a mouse model of ALS and thereby delayed the progression of the disease. This mechanism of action targeting inflammation provides a new perspective of the usage of GCSF in the treatment of ALS.

Figure 1

GCSF with sustained activity has biological efficacy and prolongs the survival of mutant SOD1 mice. Mutant SOD1 mice were treated with pegfilgrastim once a week starting at the age of 12 weeks. Plasma concentration of GCSF remained elevated multiple days after a single injection of 300 μg/kg (A) but returned back to baseline at day 7 before the next dosage (data was obtained from 33 mice, 1-2 blood samples/mouse). This occurred in a similar manner after the first dosage (A, open squares) as well as after the long-term pegfilgrastim treatment (A, black squares). Pegfilgrastim elevated the levels of stem cells (B, p <0.01, n = 11) and granulocytes (C, p <0.01, n = 11) after the first dosage, as analyzed four days after the pegfilgrastim administration, but the mobilization was decreased after prolonged pegfilgrastim treatment (n = 5). Long-term pegfilgrastim treatment increased the survival of mutant SOD1 mice as shown with a Kaplan-Meier survival graph (D, p <0.01, n = 8-9). Pegfilgrastim treatment did not delay the onset of the disease as determined by wire-hang test (E, n = 6) but increased the time from the onset to death (F, p <0.05, n = 4-5). The prolonged survival was accompanied with improved motoric performance in wire-hang test (G, p <0.05, n = 6) when analyzed within 5 weeks after the onset. The first mice died 5 weeks after the onset.

Figure 2

GCSF is neuroprotective in vitro. Primary spinal cord neuronal culture expressed >90% neurons, majority of which were interneurons, positive for NeuN (A) while a small fraction of cells were motoneurons, positive for SMI-32 (B). When spinal cord neurons were treated with GCSF for 24 h, the Akt phosphorylation (pAkt) was increased (C). β-actin is shown as a loading control and the data was normalized to β-actin control in quantification. GCSF increased pAkt as quantified from western blot data (C, p <0.05, n = 3). The molecular weight of pAkt and β-actin are 60 and 42 kDa, respectively. When spinal cord neurons were exposed to glutamate, GCSF protected the neurons from glutamate excitotoxicity (D, p <0.001, n = 18-20). GCSF did not have any effect of neuron survival in control conditions (D, n = 14). When spinal cord neuron culture was prepared from mutant SOD1 mice, we discovered that the general cell viability was slightly reduced in mutant SOD1 neurons (E, p <0.001, n = 16 embryos). GCSF increased the cell viability in mutant SOD1 neurons (E, p <0.05, n = 16 embryos). The scale bar is 50 μm.

Figure 3

GCSF with sustained activity attenuates inflammation in spinal cord in mutant SOD1 mice in vivo. Spinal cord sections were analyzed after long-term pegfilgrastim treatment from the end stage mutant SOD1 mice. The neuronal survival was decreased in mutant SOD1 mice compared to wt mice (A), determined from the ventral horn of the spinal cord. Long-term pegfilgrastim treatment did not notably rescue the cells from neurodegeneration in the spinal cord as determined with Map2, NeuN and ChAT immunostaining (3A and 3B, n = 5). However, when the inflammation status was analyzed from the same mice, long-term pegfilgrastim treatment decreased the inflammation in the spinal cord as determined with GFAP and Iba-1 immunostaining (A and B, p <0.05 and p <0.01 respectively, n = 5). The quantitation in (B) for Map2, NeuN, ChAT, GFAP and Iba-1 is shown as immunoreactive area in fold of wt mice spinal cord sections. The scale bar is 50 μm.

Figure 4

GCSF modulates inflammatory responses in vitro and ex vivo. Cells were treated in vitro with 10 ng/ml LPS for 24 h and TNFα release was quantified with ELISA. GCSF decreased the production of TNFα in primary microglia (A, p <0.001, n = 3). GCSF also decreased the TNFα production in BM monocytes obtained from wt and mutant SOD1 mice (B, p <0.001, n = 6-7). TNFα production was also decreased in BM Ly6C monocytes isolated from 20 week-old, long-term pegfilgrastim-treated mutant SOD1 mice, as compared to their vehicle-treated littermates (C, p <0.001, n = 3): TNFα production was analyzed with intracellular cytokine staining with flow cytometry after 4 h incubation of 1 μg/ml LPS ex vivo. NO production was detected with determination of NO metabolites NO₂ and NO₃ from the medium after 24 h of incubation with 10 ng/ml LPS. NO production was increased after LPS stimulation in wt (D, p <0.001, n = 4-8) and mutant SOD1 BM monocytes in vitro (D, p <0.05, n = 7-8). GCSF further increased the NO production in mutant SOD1 monocytes up to the level of wt monocytes (D, p <0.05, n = 7-8). NO production was also increased in BM (E, p <0.001, n = 3) and spleen (F, p <0.001, n = 3) leukocytes obtained from long-term pegfilgrastim-treated mutant SOD1 mice, as compared to their vehicle-treated littermates: NO production was analyzed after 24 h of incubation with 100 ng/ml LPS.

Figure 5

GCSF with sustained action increases the availability of monocytes in hematopoietic organs and monocyte recruitment into the degenerating muscle. Spleen size was decreased at late stage mutant SOD1 mice as assessed by weight (A, p <0.01, n = 6-11) and length (B, p <0.001, n = 6-11) as compared to wt control mice at the same age. Pegfilgrastim treatment increased the spleen size of mutant SOD1 mice (A, B, p <0.001, n = 6-11). The dotted line represents the spleen size of wt mice after long-term pegfilgrastim treatment (A, B). Long-term pegfilgrastim treatment also increased the number of splenocytes (C, p <0.001, n = 3) as analyzed from 20-week old mutant SOD1 mice. The number of leukocytes in BM was determined from the same mice (D): pegfilgrastim did not affect the total leukocyte number in BM but altered the composition of BM cell populations increasing the number of monocytes (E, p <0.01, n = 3). The similar effect on cell composition was detected is spleen cell populations (F, p <0.05, n = 3). The monocyte population which number was greatly increased, namely Ly6C^int had low or no production of cytokine after LPS stimulus (G, middle), in comparison to Ly6C^hi monocytes which represented high cytokine production for inflammatory stimulus (G, below) as analyzed by flow cytometry. When mononuclear cell population was analyzed from the thigh muscle including the sciatic nerve in mutant SOD1 mice at symptomatic stage, the Ly6C monocytosis was increased while lymphocytosis was decreased (H, p <0.05, n = 3).