Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jan 16;9(1):9.
doi: 10.1038/s41467-017-01881-x.

Granulocyte-colony stimulating factor controls neural and behavioral plasticity in response to cocaine

Erin S Calipari  1   2 , Arthur Godino  1   3 , Emily G Peck  1 , Marine Salery  1 , Nicholas L Mervosh  4 , Joseph A Landry  1   4 , Scott J Russo  1 , Yasmin L Hurd  1   4 , Eric J Nestler  1   4 , Drew D Kiraly  5   6   7
Affiliations

Granulocyte-colony stimulating factor controls neural and behavioral plasticity in response to cocaine

Erin S Calipari et al. Nat Commun. .

Abstract

Cocaine addiction is characterized by dysfunction in reward-related brain circuits, leading to maladaptive motivation to seek and take the drug. There are currently no clinically available pharmacotherapies to treat cocaine addiction. Through a broad screen of innate immune mediators, we identify granulocyte-colony stimulating factor (G-CSF) as a potent mediator of cocaine-induced adaptations. Here we report that G-CSF potentiates cocaine-induced increases in neural activity in the nucleus accumbens (NAc) and prefrontal cortex. In addition, G-CSF injections potentiate cocaine place preference and enhance motivation to self-administer cocaine, while not affecting responses to natural rewards. Infusion of G-CSF neutralizing antibody into NAc blocks the ability of G-CSF to modulate cocaine's behavioral effects, providing a direct link between central G-CSF action in NAc and cocaine reward. These results demonstrate that manipulating G-CSF is sufficient to alter the motivation for cocaine, but not natural rewards, providing a pharmacotherapeutic avenue to manipulate addictive behaviors without abuse potential.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing financial interests.

Figures

Fig. 1
Fig. 1
Serum multiplex analysis after self- and experimenter-administered cocaine in mice. a Timeline of experimenter-administered chronic cocaine injections. b Cocaine resulted in robust locomotor sensitization (n = 10 saline, 9 cocaine; two-way ANOVA–time: F(1,17) = 7.795, p = 0.013; cocaine treatment: F(1,17) = 326.5, p < 0.0001; interaction: F(1,17) = 9.035, p = 0.0080; p < 0.001 vs control). c Timeline of cocaine self-administration in mice (saline: n = 6 saline; cocaine: n = 10). d Average daily intake of cocaine across self-administration sessions. e Multiplex serum analysis of 32 chemokines, cytokines, and growth factors after experimenter- or self-administered cocaine. For each analyte, the heatmap depicts fold-change values compared to the respective saline group. Raw pg/ml values for each analyte and exact p values are available in Supplementary Data 1. f Correlation heatmap of individual analyte levels with either locomotor sensitization (Day 10/Day 1) or daily intake of cocaine. Exact r values for each analyte and exact p values are available in Supplementary Data 2. g G-CSF is increased after both experimenter- (two-tailed Student’s t-test; t (24) = 2.48, p = 0.020) and self-administered cocaine (t(9.40) = 2.51, p = 0.032), and G-CSF levels correlate with both h locomotor sensitization (Pearson’s r = 0.771, p = 0.025) and i daily intake of self-administered cocaine (r = 0.768, p = 0.026). j MIG is increased only after self-administered cocaine (t(10.3) = 3.74, p = 0.0036), and k, l individual MIG levels correlate only self-administered cocaine (r = 0.766, p = 0.027). m Levels of IL-1α are decreased after both experimenter-delivered (t(13.6) = 2.19, p = 0.047) and cocaine self-administration (t (13) = 7.29, p < 0.0001), however n, o IL-1α levels did not correlate with either behavior. Data represented as mean ± s.e.m. (*p < 0.05, **p < 0.01, ***p < 0.001 for Holm–Sidak post-hoc tests and t-tests)
Fig. 2
Fig. 2
G-CSF potentiates cocaine-induced neuronal activation in specific brain regions. a Experimental Timeline (left). Mice were i.p. injected with G-CSF (50 µg kg−1) or PBS 24 h and again 30 min before an injection of cocaine (20 mg kg−1 i.p.) or saline and brain tissue was collected 90 min later. c-Fos expression was measured in critical brain regions involved in the motivation to self-administer cocaine (right): G-CSF enhanced cocaine-induced neuronal activation in the mPFC and NAc. b mPFC (two-way ANOVA – cocaine: F(1,27) = 16.41, p = 0.0004; G-CSF: F(1,27) = 5.243, p = 0.030; interaction: F(1,27) = 3.759, p = 0.063), c NAc (cocaine: F(1,29) = 33.62, p < 0.0001; G-CSF: F(1,29) = 6.803, p = 0.014; interaction: F(1,29) = 5.215, p = 0.030) d While cocaine increased neuronal activation in the dorsal striatum (two-way ANOVA –cocaine: F(1,28) = 20.76, p < 0.0001), there was no added effect of G-CSF (F(1,28) = 0.05115, p = 0.82) e Similar results were observed in the ventral hippocampus (F(1,24) = 11.46, p = 0.0024; G-CSF: F(1,24) = 0.07447, p = 0.79). f c-Fos was not significantly induced by cocaine in the basolateral amygdala (F(1,31) = 2.463, p = 0.13). g c-Fos expression levels were correlated between the NAc and mPFC (Pearson’s r = 0.904, p < 0.0001). Data represented as mean ± s.e.m. (*p < 0.05, **p < 0.01, ***p < 0.001 for Holm-Sidak post-hoc tests)
Fig. 3
Fig. 3
G-CSF-related gene expression is increased after cocaine. a, b mRNA levels of Csf3 (G-CSF) in the NAc and mPFC after acute (a—two-tailed Student’s t-test – NAc: t (17) = 2.60, p = 0.019; mPFC: t (8) = 3.06, p = 0.016) and 7 days of i.p. cocaine injections (b, NAc: t (42) = 3.57, p = 0.0009; mPFC: t (27) = 1.15, p = 0.26). c mRNA levels of Csf3r (G-CSF receptor) after 7 days of i.p. cocaine injections in the NAc (t (24.1) = 2.71, p = 0.012) and mPFC (t (20) = 0.853, p = 0.40). Data represented as mean ± s.e.m. (*p < 0.05, ***p < 0.001 for t-tests)
Fig. 4
Fig. 4
Detection of G-CSF and G-CSFR in the NAc of D1-tdTomato mice. Immunolabeling for tdTomato protein and GCSFR or GCSF in D1-tdTomato mice was performed to determine cell-type expression in the NAc. a Representative confocal images acquired in the shell of the NAc from control animals, demonstrating expression of G-CSFR (upper) and G-CSF (lower) in multiple cell types. b Representative images from mice treated with cocaine (20 mg kg−1, i.p. × 7 days) again showing expression of G-CSFR (upper) and G-CSF (lower). For all images nuclei were counterstained with DAPI, tdTomato protein is labeled in red, G-CSF and G-CSFR are labeled in green. Scale bar = 20 µm
Fig. 5
Fig. 5
G-CSF levels are increased by the selective activation of mPFC to NAc projections. a Experimental design of projection-specific DREADD stimulation. Mice were injected with a retrograde traveling CAV2-Cre virus in the NAc and a Cre-dependent hM3Dq-DREADD virus in either the mPFC or the VTA to allow for the specific stimulation of either mPFC to NAc or VTA to NAc. b Csf3 (G-CSF) mRNA levels in the NAc were increased after mPFC to NAc stimulation (one-way ANOVA; F(2,12) = 13.4, p = 0.0009, Sidak post-hoc: p = 0.0037 vs control). c Csf3r (G-CSFR) mRNA levels in the NAc were increased only after mPFC to NAc stimulation (F(2,12) = 8.14, p = 0.0058, p = 0.0093 vs control). d Peripheral G-CSF serum levels were not affected by stimulation (F(2,12) = 1.82, p = 0.20). e Mice were injected i.p. for 7 days with cocaine methiodide (CocMet), a cocaine analog that does not cross the blood brain barrier, to assess the effects of peripheral cocaine on G-CSF. Cocaine methiodide chronic treatment had no effect on f Csf3 (G-CSF) mRNA levels in the NAc (two-tailed Student’s t-test; t (12) = 0.772, p = 0.45), g Csf3r (G-CSFR) mRNA levels in the NAc (t (12) = 1.11, p = 0.29), or h G-CSF serum levels (t (12) = 0.631, p = 0.54). Data represented as mean ± s.e.m. (**p < 0.01 for Sidak post-hoc tests)
Fig. 6
Fig. 6
G-CSF enhances cocaine-induced locomotor sensitization and CPP. (a—top) Experimental timeline of locomotor sensitization to cocaine. Mice (n = 4 PBS, 5 G-CSF) were i.p. injected with G-CSF (50 µg kg−1) or PBS 1 h before monitoring locomotor activity following an injection of saline or cocaine (7.5 mg kg−1). (a—bottom) Locomotor sensitization to cocaine was increased in mice pre-treated with G-CSF (repeated-measures two-way ANOVA; time: F (6,42) = 33.16, p < 0.0001; G-CSF: F(1,7) = 8.808, p = 0.021; interaction: F(6,42) = 3.942; p = 0.0032). b For cocaine conditioned place preference, mice were injected with G-CSF (50 µg kg−1) or PBS 1 h every day before testing. Two-way ANOVA testing demonstrated a main effect of G-CSF (F(1,36) = 11.76, p = 0.0015), and Holm-Sidak post-hoc testing demonstrated increased CPP in G-CSF-treated mice conditioned with 3.75 mg kg−1 of cocaine (PBS n = 6; G-CSF: n = 9; p < 0.05 vs PBS), 7.5 mg kg−1 (PBS: n = 6; G-CSF: n = 10; p < 0.05 vs PBS) of cocaine but not with 15 mg kg−1 (PBS: n = 5; G-CSF: n = 6). (*p < 0.05, **p < 0.01 for Holm-Sidak post-hoc tests; # p < 0.05, ## p < 0.01 for two-way ANOVA main effects)
Fig. 7
Fig. 7
Neutralization of central G-CSF signaling reduces conditioned place preference. a To determine the role of circulating G-CSF in behavior, mice (IgG: n = 6; α-G-CSF antibody: n = 9) were i.p. injected with anti-G-CSF antibody (10 µg) or pre-immune IgG control antibody 1 h every day before testing for CPP at 7.5 mg kg−1 of cocaine. Systemic anti-G-CSF antibody did not significantly affect cocaine CPP (Two-tailed Student’s t-test: t (13) = 1.48, p = 0.16). b To test the effects of blocking signaling in the NAc anti-G-CSF antibody or pre-immune IgG (1 µg/side) was infused into the NAc via continuous osmotic minipump before and during testing for CPP at 7.5 mg kg−1 of cocaine (IgG: n = 5; α-G-CSF antibody: n = 10). NAc infusion of anti-G-CSF antibody blocked cocaine CPP (t (12.4) = 3.75, p = 0.0026). Data represented as mean ± s.e.m (**p < 0.01 for t-test)
Fig. 8
Fig. 8
G-CSF increases the motivation to self-administer cocaine. a Experimental timeline. Rats were trained to self-administer cocaine on a fixed-ratio one (FR1) schedule of reinforcement. Animals went through two behavioral tasks: the threshold procedure to assess motivation, and FR1 self-administration to assess consumption. b Acquisition for the two experimental groups before treatment showing that there were no significant differences before the experiment (two-way ANOVA: time: F(4,64) = 69.11, p < 0.0001; group: F(1,16) = 0.00147, p = 0.97; interaction: F(4,64) = 0.1985, p = 0.94). c Cocaine intake did not differ in the two groups before G-CSF or PBS treatment (Student’s t-test; t (16) = 0.427, p = 0.34). d Dose-response curves. G-CSF pretreatment increased responding for lower doses of cocaine indicating an upward shift in the dose–response function (two-way ANOVA; price: F(9,135) = 42.06, p < 0.0001; G-CSF: F(1,15) = 4.623, p < 0.05; interaction: F(9,135) = 2.616, p < 0.01). e, f Representative demand curves, plotting consumption of cocaine as a function of price, from a PBS-treated control and G-CSF-treated animal. g Averaged demand curves from both groups (two-way ANOVA; price: F(9,135) = 72.88, p < 0.0001; G-CSF: F(1,15) = 5.036, p = 0.40; interaction: F(9,135) = 0.9677, p = 0.47) h Pmax is increased in animals treated with G-CSF (Student’s t-test; t(14) = 2.002, *p = 0.033). i Q 0 levels are increased in G-CSF treated animals (t (14) = 2.374, *p = 0.016). j Intake was also measured using a fixed ratio one schedule of reinforcement. Animals treated with G-CSF took more cocaine injections in a three-hour session than their PBS-treated counterparts (Student’s t-test for active lever presses; t = 2.866, df = 15, p = 0.012). Data represented as mean ± s.e.m (*p < 0.05, **p < 0.01, p < 0.001 for Holm–Sidak post-hoc tests and t-tests; # p < 0.05 for two-way ANOVA main effects)
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Siciliano CA, Fordahl SC, Jones SR. Cocaine self-administration produces long-lasting alterations in dopamine transporter responses to cocaine. J. Neurosci. 2016;36:7807–7816. doi: 10.1523/JNEUROSCI.4652-15.2016. - DOI - PMC - PubMed
    1. Wolf ME. Synaptic mechanisms underlying persistent cocaine craving. Nat. Rev. Neurosci. 2016;17:351–365. doi: 10.1038/nrn.2016.39. - DOI - PMC - PubMed
    1. Robison AJ, Nestler EJ. Transcriptional and epigenetic mechanisms of addiction. Nat. Rev. Neurosci. 2011;12:623–637. doi: 10.1038/nrn3111. - DOI - PMC - PubMed
    1. Castells X, Cunill R, Pérez-Mañá C, Vidal X, Capellà D. Psychostimulant drugs for cocaine dependence. Cochrane Database Syst. Rev. 2016;9:CD007380. - PMC - PubMed
    1. Araos P, et al. Plasma profile of pro-inflammatory cytokines and chemokines in cocaine users under outpatient treatment: influence of cocaine symptom severity and psychiatric co-morbidity. Addict. Biol. 2015;20:756–772. doi: 10.1111/adb.12156. - DOI - PubMed

Publication types

MeSH terms

Substances