Mechanism of Brain Changes in Addiction

1521-0081/67/4/8721004$25.00 http://dx.doi.org/10.1124/pr.115.010967PHARMACOLOGICAL REVIEWS Pharmacol Rev 67:8721004, October 2015Copyright 2015 by The American Society for Pharmacology and Experimental Therapeutics

ASSOCIATE EDITOR: MARKKU KOULU

Mechanisms of Action and Persistent Neuroplasticityby Drugs of Abuse

Esa R. Korpi, Bjrnar den Hollander, Usman Farooq, Elena Vashchinkina, Ramamoorthy Rajkumar, David J. Nutt, Petri Hyyti,and Gavin S. Dawe

Department of Pharmacology, Faculty of Medicine, University of Helsinki, Finland (E.R.K., B.d.H., E.V., P.H.); Department ofPharmacology, Yong Loo Lin School of Medicine, National University Health System, Neurobiology and Ageing Programme, Life SciencesInstitute, National University of Singapore, Singapore, and SINAPSE, Singapore Institute for Neurotechnology, Singapore (E.R.K., R.R.,

G.S.D.); Interdepartmental Neuroscience Program, Yale University, New Haven, Connecticut (U.F.); and Centre forNeuropsychopharmacology, Division of Brain Sciences, Burlington Danes Building, Imperial College London, London.

United Kingdom (D.J.N.)

Abstract. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875I. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 875

A. Different Forms of Neuroplasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 876B. Short-term Neuroplasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 878C. Main Forms of Long-term Neuroplasticity: Long-term Potentiation and Long-term

Depression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8791. Presynaptic Forms of Long-term Plasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8802. Postsynaptic Forms of Long-term Plasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8803. a-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid and N-methyl-D-aspartate

Receptor Phosphorylation.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8804. a-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid and N-methyl-D-aspartate

Receptor Trafficking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881D. Forms of Long-term Plasticity at Inhibitory Synapses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881E. Developmental Maturation of the Brain in Rodents and Humans . . . . . . . . . . . . . . . . . . . . . . . . . 882F. Developmental Neuroplasticity: Critical Periods, Reopening of Plasticity in Adults . . . . . . . . 883G. Neurotrophins as Regulators of Neuroplasticity: Brain-Derived Neurotrophic

Factor as an Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885H. Animal Models of Drug Reinforcement and Addiction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 885

II. Actions and Persistent Effects of Specific Drugs of Abuse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886A. Cocaine, a Stimulant and Local Anesthetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886

1. Persistent Ventral Tegmental Area Neuroplasticity after a Single Dose. . . . . . . . . . . . . . . 8872. Changes in the Nucleus Accumbens Mediate Cocaine Addiction and Relapse.. . . . . . . . . 8903. Altered Gene Expression in the Nucleus Accumbens in Cocaine Seeking and Relapse. 8934. Progressive Involvement of the Dorsal Striatum.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8955. Changes in Lateral Habenula and Rostromedial Tegmental Nucleus Contribute

to Aversive Symptoms of Cocaine Withdrawal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8956. Complex Alterations in the PFC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8967. Hippocampal Functional Changes.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8978. Bed Nucleus of Stria Terminalis and Amygdala are Involved in

Stress-Induced Reinstatement of Cocaine Seeking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8979. Effects of Adolescent Cocaine Exposure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89810. Human Imaging Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89811. Limited Evidence of Cocaine Neurotoxicity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 899

This project was partially funded by the Academy of Finland, the Sigrid Juselius foundation, the Finnish Foundation for Alcohol Studies,the Jane and Aatos Erkko Foundation, the Foundations Professor pool, and the Orion Research Foundation.

Address correspondence to: Dr. Esa R. Korpi, Department of Pharmacology, Faculty of Medicine, Biomedicum Helsinki, POB 63(Haartmaninkatu 8), FI-00014 University of Helsinki, Finland. E-mail [email protected] or Gavin S. Dawe, Department of Pharmacology,Yong Loo Lin School of Medicine, Building MD3, #04-01Y, 16 Medical Drive, Singapore 117600. Email: [email protected].

dx.doi.org/10.1124/pr.115.010967

872

at Nyu Med Centr Lib on O

ctober 8, 2015pharm

rev.aspetjournals.orgD

ownloaded from

12. Potential Mechanisms in Cocaine Neurotoxicity and Neuroprotection. . . . . . . . . . . . . . . . . 90013. Methylphenidate and Novel Psychoactive Substances.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90014. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 900

B. Amphetamine-type Psychostimulants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9001. Many Stimulant Amphetamines for Abuse and Therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9022. Molecular Targets and Mechanisms of Action of Amphetamines. . . . . . . . . . . . . . . . . . . . . . 9023. Comparison of the Mechanisms of Action of Amphetamine versus Methamphetamine. 9024. Effects of Amphetamine at Plasmalemmal Dopamine Transporter. . . . . . . . . . . . . . . . . . . . 9035. Regulation of Dopamine Efflux Via the Dopamine Transporter by Protein Kinase

C and Ca2+/Calmodulin Kinase II Phosphorylation, and by Reactive Oxygen Species. . 9046. Effects on Secretory Vesicles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9047. Other Amphetamine Targets: Monoamine Oxidase, Tyrosine Hydroxylase, other Mono-

amine Transporters, Trace Amine-Associated Receptor 1, and Sigma Receptors. . . . . . . 9058. Action of Substituted Amphetamines and Cathinones. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9059. Action of Stimulants Common in Clinical Use. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90610. Efficacy and Addiction Liability of Amphetamines in Clinical Use. . . . . . . . . . . . . . . . . . . . 90611. Neuroplasticity Related to Sensitization and Addiction to Amphetamines. . . . . . . . . . . . . 90712. Glutamate and Amphetamine/Methamphetamine Reinforcement/Extinction/Reinstatement.. 90813. Long-term Neuroadaptation/Neurotoxicity after Exposure to High Doses of

Psychostimulants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909a. Effects on brain structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909b. Effects on neurochemistry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 910c. Effects on behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 910

14. Limited Evidence for Long-term Effects of Substituted Cathinones in RodentModels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911

15. Mechanisms Involved in Long-term Adaptation and Neurotoxicity. . . . . . . . . . . . . . . . . . . . 911a. Oxidative stress: DA oxidation and drug metabolites as sources of reactive species.911b. Mitochondrial dysfunction: ATP deficit, superoxide leakage, and apoptosis.. . . . . . . . 912

ABBREVIATIONS: AC, adenylyl cyclase; ADHD, attention deficit hyperactivity disorder; 2-AG, 2-arachidonoylglycerol; ALDH,aldehyde dehydrogenase; AMPA, a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; AMPH, amphetamine; BDNF, brain-derivedneurotrophic factor; BLA, basolateral complex of the amygdala; BNST, bed nucleus of stria terminalis; BOLD, blood oxygen leveldependent; BZ, benzodiazepine; CAM, cell adhesion molecule; CaMKII, Ca2+/calmodulin kinase II; CCK, cholecystokinin; CeA, centralnucleus of the amygdala; CNS, central nervous system; CPP, conditioned place preference; CREB, cAMP response element-bindingprotein; CRF, corticotropin-releasing factor; D1R, dopamine-1 receptor; D2R, dopamine-2 receptor; DA, dopamine; DARPP-32, DA- andcAMP-regulated phosphoprotein; DAT, dopamine transporter; DBI, diazepem binding inhibitor; DGL, diacylglycerol lipase; DOI,2,5-dimethoxy-4-iodoamphetamine; DSE, depolarization-induced suppression of excitation; DSI, depolarization-induced suppression ofinhibition; eCB, endocannabinoid; ECM, extracellular matrix; EEG, electroencephalography; EPSC, excitatory postsynaptic currents;ER, endoplasmic reticulum; ERK, extracellular signal-regulated kinase; ETC, electron transport chain; FC, frontal cortex; fMRI,functional magnetic resonance imaging; G9a, lysine dimethyltransferase G9a; GABA, g-aminobutyric acid; GHB, gamma-hydroxybutyrate; GPCR, G protein-coupled receptor; HC, hippocampus; HDAC, histone deacetylase; HFS, high-frequency stimulation;HPPD, hallucinogen persisting perception disorder; 5-HT, 5-hydroxytryptamine, serotonin; ICSS, intracranial self-stimulation; iGluR,iontropic glutamate receptor; IL, infralimbic cortical region; IP3, inositol 1, 4,5-trisphosphate; IPN, interpeduncular nucleus; IPSC,inhibitory postsynaptic currents; IPSP, inhibitory postsynaptic potential; JNK, c-Jun N-terminal kinases; Kal-7, postsynaptic density-localized kalirin-7; KCC, potassium-chloride cotransporter; KO, knockout; LA, lateral nucleus of amygdala; LC, locus ceruleus; LDTg,laterodorsal tegmental nucleus; LDX, lisdexamfetamine; LFS, low-frequency stimulation; LHb, lateral habenula; LSD, lysergic aciddiethylamide; LTD, long-term depression; LTP, long-term potentiation; MAO, monoamine oxidase; MAPK, mitogen-activated proteinkinase; MDMA, 3,4-methylenedioxymethamphetamine (ecstasy); MDMC, methylone; METH, methamphetamine; mGlu, metabotropicglutamate receptor; MHb, medial habenula; MK-801, dizocilpine; 4-MMC, mephedrone; MPH, methylphenidate; MRI, magneticresonance imaging; MSN, medium spiny neuron; mTOR, mammalian target of rapamycin; NAc, nucleus accumbens; nAChR, nicotinicacetylcholine receptors; NE, norepinephrine; NET, norepinephrine transporter; NFkB, nuclear factor kappa-B; NMDA, N-methyl-D-aspartate; NO, nitric oxide; NOP, nociceptin receptor; N/OFQ, nociceptin/orphanin FQ; NPY, neuropeptide Y; OFC, orbitofrontal cortex;OR, odds ratio; OX, orexin; P, postnatal day; PAG, periaqueductal gray area; PBR, peripheral benzodiazepine receptor; PCP,phencyclidine; PET, positron emission tomography; PFC, prefrontal cortex; PI3K, phosphatidylinositide 3-kinase; PKA, protein kinaseA; PKC, protein kinase C; PKMzeta, atypical PKC; PLC, phospholipase C; PNN, perineuronal nets; PrL, prelimbic cortical region; PSA-NCAM, polysialylated neuronal cell adhesion molecule; PSD, postsynaptic density; Pv, parvalbumin; R, receptor; RGS, regulator of Gprotein signaling; rmTg, rostromedial tegmental nucleus; ROS, reactive oxygen species; SA, self-administration; SERT, serotonintransporter; SN, substantia nigra; TA1, trace amine-associated receptor 1; TH, tyrosine hydroxylase; THC, tetrahydrocannabinol;TLR4, toll-like receptor 4; TPH, tryptophan hydroxylase; TrkB, tropomyosin receptor kinase B; TSPO, mitochondrial translocatorprotein; VGCC, voltage-gated calcium channel; vHC, ventral hippocampus; VMAT-2, vesicular monoamine transporter 2; VTA, ventraltegmental area.

Drug-Induced Neuroplasticity 873

c. Excitotoxicity: glutamate receptor involvement and intracellular excitotoxicprocesses.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913

d. Hyperthermia: a catalyst of other toxic processes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 913e. Other factors: inflammation, blood-brain barrier disruption, protective

preconditioning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91416. Evidence of Long-term Plasticity and Cognitive Effects after Stimulant Use in

Humans.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914a. Cognitive behaviors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914b. Brain imaging of receptors, transporters and activity/metabolism.. . . . . . . . . . . . . . . . . 915c. Magnetic resonance imaging studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915d. Magnetic resonance spectroscopy and analysis of postmortem tissue. . . . . . . . . . . . . . . 916e. Prospective and experimental studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916

17. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917C. Nicotine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917

1. Nicotinic Acetylcholine Receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9172. Upregulation of Nicotinic Receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9183. Nicotine-priming Effects and Upregulation of Catecholamine Synthesis. . . . . . . . . . . . . . . 9194. Nicotine and Neuroplasticity within the Ventral Tegmental Area. . . . . . . . . . . . . . . . . . . . . 9205. Nicotine and Neuroplasticity in the Lateral Hypothalamic Orexin System. . . . . . . . . . . . 9216. Nicotine and Glutamatergic Neuroplasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9217. Nicotine and the Habenula. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9228. Long-term Sequelae of Adolescent Exposure to Nicotine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9239. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 924

D. Neural Adaptations Induced by Ethanol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9251. Cell Membrane Ion Channels as Primary Targets of Ethanol. . . . . . . . . . . . . . . . . . . . . . . . . 9252. Ethanol-Induced Changes in Glutamatergic Transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . 9283. Ethanol-Induced Changes in g-Aminobutyric Acidergic Transmission. . . . . . . . . . . . . . . . . 9294. Ethanol-Induced Changes in Neuropeptide Mechanisms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9315. Role of Acetaldehyde in Ethanols Reinforcing Actions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9336. Structural Plasticity in Alcohol Dependence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9337. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934

E. Benzodiazepines and Other GABAergic Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9341. Molecular Targets for Benzodiazepines.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9352. Neuroplasticity Induced by Benzodiazepines and Related Compounds. . . . . . . . . . . . . . . . 9363. Behavioral After-effects of g-Aminobutyric Acid A Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9394. Effects of Flumazenil on Benzodiazepine and Alcohol Tolerance. . . . . . . . . . . . . . . . . . . . . . 9395. Treatment of Addictions by g-Aminobutyric Acid B Receptor Agonists. . . . . . . . . . . . . . . . 9406. g-Hydroxybutyrate as a Drug of Abuse and a Therapeutic Compound.. . . . . . . . . . . . . . . . 9417. Anesthetics and Neuroplasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9428. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943

F. N-methyl-D-aspartate Receptor Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9431. Ketamine: a Dissociative Anesthetic with Rapid Antidepressant Effects in Patients. . . 9442. Phencyclidine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9463. Dizocilpine, a Prototypic Noncompetitive N-methyl-D-aspartate Receptor Antagonist. . 9484. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949

G. Opioid-Induced Neural Adaptations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9491. Desensitization and Internalization in Opioid Tolerance and Dependence. . . . . . . . . . . . . 9502. Cellular Signaling in Opioid Tolerance and Dependence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9503. Novel Mechanisms in Between-systems Adaptations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9514. Opioid-Induced Changes in Excitatory and Inhibitory Synaptic Plasticity. . . . . . . . . . . . . 9525. Chronic Opioids and Synaptic Plasticity.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9536. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 955

H. Cannabinoids: Multiple Mechanisms and Possible Indications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9551. Endocannabinoid System as an Endogenous Lipid Messenger System.. . . . . . . . . . . . . . . . 9562. Short-term Plasticity Involving Retrograde Endocannabinoid Signaling. . . . . . . . . . . . . . . 9573. Long-term Plasticity and CB1 Receptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9584. Cannabinoid Effects on Brain Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 958

874 Korpi et al.

5. Cannabinoid-Induced Cognitive Impairment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9596. Rewarding and Aversive Behaviors, Specific Actions on VTA DA Neurons. . . . . . . . . . . . 9607. Cannabinoids during Adolescence: Increased Risk for Schizophrenia?. . . . . . . . . . . . . . . . . 9618. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962

I. Hallucinogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9621. Effects of Serotonergic Hallucinogens Are Mediated by the 5-HT2A Receptor. . . . . . . . . . 9622. Hallucinogen ActionA Perturbation of Sensory Gating or Desynchronization

of Cortical Rhythms?.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9633. Behavioral Effects and Addiction Potential. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9644. Long-term Residual Effects of Serotonergic Hallucinogens: the Role of

Neuroplasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 964a. Sensory processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 964b. Mood and anxiety. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965

5. Scopolamine: Another Hallucinogen Revisited for Depression. . . . . . . . . . . . . . . . . . . . . . . . . 9666. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 966

III. General Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 966A. Novel Methods for Future Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 967References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 969

AbstractAdaptation of the nervous system todifferent chemical and physiologic conditions is impor-tant for the homeostasis of brain processes and forlearning and remembering appropriate responses tochallenges. Although processes such as tolerance anddependence to various drugs of abuse have been knownfor a long time, it was recently discovered that evena single pharmacologically relevant dose of variousdrugs of abuse induces neuroplasticity in selectedneuronal populations, such as the dopamine neuronsof the ventral tegmental area, which persist long afterthe drug has been excreted. Prolonged (self-) adminis-tration of drugs induces gene expression, neurochem-ical, neurophysiological, and structural changes inmany brain cell populations. These region-specificchanges correlate with addiction, drug intake, and

conditioned drugs effects, such as cue- or stress-induced reinstatement of drug seeking. In rodents,adolescent drug exposure often causes significantlymore behavioral changes later in adulthood thana corresponding exposure in adults. Clinically themost impairing and devastating effects on the brainare produced by alcohol during fetal development. Inadult recreational drug users or in medicated patients,it has been difficult to find persistent functional orbehavioral changes, suggesting that heavy exposure todrugs of abuse is needed for neurotoxicity and forpersistent emotional and cognitive alterations. Thisreview describes recent advances in this importantarea of research, which harbors the aim of translatingthis knowledge to better treatments for addictions andrelated neuropsychiatric illnesses.

I. Introduction

Brain diseases are associated with an enormous costto affected individuals, their families, and the society. InEurope, it has been estimated that the total yearly costof brain diseases in 2010 was close to 800 billion euros(Olesen et al., 2012), of which addictions are responsiblefor as much as anxiety disorders, with only dementiaand mood illnesses costing more (DiLuca and Olesen,2014). Drug abuse produces both direct and indirectcosts to the society, although many of the drugs are alsoclinically used to treat various patient groups. Thepurpose of this review is to present up-to-date knowl-edge of the mechanisms of action of the main drugs ofabuse and to reveal the possible long-term alterations inthe nervous system associated with the use and abuse ofvarious drugs acting on the brain, also paying attentionto the trajectory of brain development.Addiction is a complex phenomenon, which is not only

dependent on pharmacological mechanisms, but alsohas a societal/cultural dimension. This is reflected inthe proportion of drug addiction among different age

and ethnic groups studied in the United States. Indi-viduals with alcohol or illicit drug abuse or dependencein the past year constituted 5% of the adolescents agedbetween 12 and 17 and more than 8% of all individualsaged 12 or older (SAMSHA, 2014a). Of the racial/ethnicgroups studied among those aged 12 and older, Asianshad the lowest proportion of binge alcohol drinkers andpast-month illicit drug users, likely partly due tosocietal/cultural traditions. These findings on ratherwidespread drug and alcohol exposures at young agesare very alarming, because brain development contin-ues well past the age of 17 years and because theexposure to several different drugs, such as illicit drugs,cigarettes, and alcohol, appears to concentrate on thesame individuals (SAMSHA, 2014b). Therefore, it ispossible that harmful effects from early drug use willprevail later in life, because drugs of abuse inducedifferent modulations in brain circuitries (adaptation,plasticity, learning, and memory) due to their pharma-cological actions and due to behavioral/social effectsassociated with their use and settings. Thus, we will


explore the data on different drugs and how theypersistently affect brain functions if the drug exposureoccurs during adolescence.The review first gives a general basic introduction to

neuroplasticity. Then sections on different drugs followand they have different focuses, because the variousdrugs do not invoke exactly similar mechanisms oradaptations in the nervous system. The review will endwith a short general summary.Certain mechanisms seem to be common for many

drugs, for example, activation of the extracellularsignal-regulated kinase (ERK) pathway in specific brainstructures is necessary for effects of and tolerance tococaine, nicotine, MDMA, phencyclidine, alcohol, andcannabinoids after both acute and chronic treatments inrodents (Kyosseva et al., 2001; Salzmann et al., 2003;Valjent et al., 2004; Rubino et al., 2005; Tonini et al.,2006; Schroeder et al., 2008). One human postmortembrain study suggested that cocaine, cannabis, and/orphencyclidine abuse all decrease transcription ofcalmodulin-related genes and increase transcription ofgenes related to lipid/cholesterol and Golgi/endoplasmicreticulum (ER) function in the anterior prefrontal cortex(PFC), whichmay underlie changes in synaptic functionand plasticity (Lehrmann et al., 2006). However, post-mortem brain regional gene expression profiling inalcohol-dependent patients has indicated widely differ-ing sets of affected genes between different brainregions (Flatscher-Bader et al., 2005, 2006), with alco-holics nevertheless being easily separable from non-alcoholic controls and smokers (Flatscher-Bader et al.,2010). The same situation was found for other drugs ofabuse (Albertson et al., 2004, 2006). Significant alter-ations in glutamate and GABA receptor mRNAs werefound in postmortem brains of alcohol-dependent sub-jects, but the changes differed from one brain region toanother (Jin et al., 2011, 2014a,b; Bhandage et al.,2014). For these obvious reasons, we reviewed theliterature for each drug according to the specific brainregions where alterations have been observed mostcommonly in preclinical experiments. These brain areasand pathways, which serve for positive and negativereinforcing behavioral and emotional effects and forgoal-directed and habitual drug seeking, are illustratedin Fig. 1. However, it will be necessary in the future tostudy the plasticity and mechanisms more preciselyat the level of different neuronal populations andsubpopulations.As is usual in addiction-related reviews and research,

the dopamine (DA) mechanisms are center stage. Thereader is referred to recent reviews on various aspectsof DA as a neurotransmitter and a regulator of motor,cognitive, andmotivatedbehavior (BjorklundandDunnett,2007; Beaulieu and Gainetdinov, 2011; Salamone andCorrea, 2012). DA is important for a wide number ofbrain functions, and it will be impossible to cover eachand every aspect of the DA mechanisms in drug actions

and adaptations. The focus in this review will be moreon the neuroplasticity of the glutamate synapses on DAneurons.Pharmacological actions of drugs are mediated by

specific receptors. On the other hand, the effects of thedrugs of abuse are often modulated by experimentalsettings and expectancies in both preclinical studiesand human experiments. Drug intake in rodents isdependent, for example, on cage conditions, with theeffects differing between opioids and stimulants (Badianiet al., 2011). Voluntary self-administration (SA) (self-stimulation or cocaine) induces different brain regionalactivations (Porrino et al., 1984), or more prolongedsynaptic molecular adaptations, than experimenter-given stimulation or drug injections (Chen et al., 2008).Placebo/nocebo effects are real in human studies, and theexpectancy of strong or uncertain drug effects are knownto affect human brain imaging results in response toacute drugs (Volkow et al., 2010). For example, insmokers, positive and negative beliefs on nicotine contentof cigarettes influenced functional magnetic resonanceimaging (fMRI)-scanned striatal responses to value andreward prediction errors during an investment task (Guet al., 2015), indicating that beliefs can affect cognitiveperformance also under the drug-associated states. Theseissues indicate that, in addition to direct pharmacologicalactions, abused drugsmay also change basic learning andmemory processes, for example, by conditioning andenvironmental factors.

A. Different Forms of Neuroplasticity

The ability of the brain to remodel its connectionsfunctionally and structurally in response to individualexperience has been described by the concept of neuro-plasticity. Neuroplasticity occurs on a variety of levelsranging from molecular changes in synapses to large-scale changes involved in neurocircuitry remapping.Synaptic plasticity refers to adaptive changes in thestrength of synaptic connections. On the basis of its timeframe, synaptic plasticity has been classified as short-term (acts on a timescale of milliseconds to minutes)and long-term (hours to days) plasticity. Short-termplasticity is achieved through transient changes, suchas facilitation or depression of a synaptic connection,which then quickly return to their initial state. How-ever, repeated stimulation causes a persistent changein the connection to achieve long-term plasticity.Hebbian or activity-dependent plasticity is the most

studied form of long-term plasticity. It occurs whenpresynaptic stimulation coincides with postsynapticdepolarization (Hebb, 1949; Bi and Poo, 2001). Thebest-known example of Hebbian plasticity is N-methyl-D-aspartate receptor (NMDAR)-mediated long-term po-tentiation (LTP). Importantly, this form of LTP occursonly in synapses that actively contribute to the in-duction process, so it is input and synapse specific. Theterm anti-Hebbian plasticity currently describes

876 Korpi et al.

either long-term depression (LTD) (Nelson, 2004) or LTPthat occurs when presynaptic activation coincides withpostsynaptic inactivity (Kullmann and Lamsa, 2007).In rodents, the neuronal organization remains imma-

ture at birth (see below). The process that describeschanges in neuronal organization during develop-ment as a result of environmental interactions andexperience/learning-induced neural changes is knownas developmental plasticity. To maintain the balancebetween neuronal excitation and inhibition, homeo-static plasticity regulates the overall activity of complexcircuits by specifically regulating the destabilizing

effects of developmental and learning processes. Other-wise, activity-dependent forms of plasticity could driveneural activity toward runaway excitation or quies-cence (Miller, 1996; Turrigiano, 1999; Turrigiano andNelson, 2004).Metaplasticity refers to plasticity of synaptic plas-

ticity, which describes changes in the ability to inducefurther synaptic plasticity (Abraham and Bear, 1996).For example, prolonged exposure to cocaine inducesa population of silent glutamatergic synapses in thenucleus accumbens (NAc) that form sites for futureplasticity (reviewed in Lee and Dong, 2011).

Fig. 1. Brain regions of major importance for the acute and chronic addictive effects of drugs of abuse. The top drawing shows many pathways thatparticipate in rewarding/drug-seeking behavior. The middle drawing depicts several pathways/brain regions that are particularly related to addiction-related aversive behaviors (wide connections of the locus ceruleus to other brain areas are not shown). The bottom drawing illustrates the role of midbrain-striatal-cortical loops in goal-directed and habitual drug taking. See Fig. 10 for additional pathways involved in nicotine neurotoxicity. Amy, amygdala;BNST, bed nucleus of stria terminalis; CeA, central nucleus of amygdala; DLS, dorsolateral striatum; DMS, dorsomedial striatum; HC, hippocampus; LC,locus ceruleus; LDTg, laterodorsal tegmental nucleus; LH, lateral hypothalamus; LHb, lateral habenula; mPFC, medial prefrontal cortex; NAc, nucleusaccumbens; OFC, orbitofrontal cortex; rmTg, rostromedullary tegmental nucleus; SN, substantia nigra; VTA, ventral tegmental area.


To help appreciate the effects of abused drugs onsynaptic plasticity, in the following paragraphs we willbriefly introduce mechanisms of the most commonpresynaptic and postsynaptic forms of synapticplasticity.

B. Short-term Neuroplasticity

Short-term plasticity can appear as a transient facil-itation, depression, or augmentation and posttetanicpotentiation in synaptic strength that lasts for up toa few minutes (reviewed in Fioravante and Regehr,2011). Despite the variety in synaptic neurotransmit-ters, all forms of short-term plasticity are primarilygoverned by presynaptic mechanisms associated withfluctuations of presynaptic residual Ca2+, which acts onone or more molecular targets, resulting in the changesin neurotransmitter release (Fig. 2). Enhancement of

Ca2+ channel activity and increases in the probability ofCa2+ influx, altered vesicle pool properties, local de-pletion of Ca2+ buffers, and increases in quantal size ofneurotransmitter release contribute to short-term fa-cilitation (Fioravante and Regehr, 2011). On the otherhand, vesicle depletion, inactivation of neurotransmit-ter release sites, and Ca2+ channels contribute to short-term synaptic depression (Neher and Sakaba, 2008). Inaddition, glial-neuronal interactions impact on short-term synaptic plasticity by controlling the speed andextent of neurotransmitter clearance from the synapticcleft (Bergles et al., 1999) as well as by astroglial releaseof substances that can affect synaptic efficacy (Araqueet al., 2001). Most importantly, there is retrogradecommunication between post- and presynaptic termi-nals: both endocannabinoids (reviewed in Wilsonand Nicoll, 2002; Kano et al., 2009) and nitric oxide

Fig. 2. Typical induction protocols and main factors regulating mechanisms of short-term plasticity. (A) Brief paired-pulse (PP) stimulation inducesshort-term facilitation of neurotransmission by transiently increasing the Ca2+-dependent readily releasable pool (RRP) of synaptic vesicles. (B) Short-term depression of neurotransmission can be induced by frequent tetanic stimulation, which transiently depletes synaptic vesicles. (C). Depolarization-induced suppression of inhibition (DSI) or similarly that of excitation (DSE; not shown) is a locally induced transient depression of neurotransmissionthat is dependent on retrograde eCB signaling. Depolarization-induced suppression of inhibition/excitation begins with stimulation of excitatoryneuronal connections inducing eCB synthesis, and 2-AG in particular then moves to activate presynaptic CB1 receptors on surrounding neurons. Thisinduces local, transient depression of inhibitory or excitatory (not shown) neurotransmission.

878 Korpi et al.

(NO)-guanylyl cyclase signaling (Sammut et al., 2010)contribute in a general transient local process thattunes neurotransmitter release and different aspects ofsynaptic dynamics both in inhibitory and excitatorysynapses (discussed in more detail in sections II.G andII.H). Short depolarization of a neuron may causea transient suppression of excitation or inhibition ofthat and neighboring neurons, called depolarization-induced suppression of excitation or inhibition (DSE

and DSI), which was found to be dependent on retro-grade encocannabinoid (eCB) signaling (see section II.Hon cannabinoids).

C. Main Forms of Long-term Neuroplasticity:Long-term Potentiation and Long-term Depression

Long-term forms of synaptic plasticity can appear aspotentiation (long-term potentiation, LTP) or depres-sion (long-term depression, LTD) in synaptic strength

Fig. 3. Main factors regulating pre- and postsynaptic plasticity mechanisms, such as LTP and LTD, after most typical induction protocols. (A)Presynaptic LTP is often measured using paired-pulse (PP) stimulation e.g., at 50-ms interstimulus intervals (ISI). This leads to presynaptic Ca2+

influx via VGCCs, activation of adenylate cyclase (AC) and phosphorylation of synaptic vesicular proteins, such as Rab3a and RIM1a, leading toincreased transmitter release. (B) During low-frequency stimulation (LFS; often used for LTD induction) weak presynaptic depolarization triggers onlya modest Ca2+ influx that activates phosphatases (calcineurin, protein phosphatase 1, and protein phosphatase 2) in the postsynaptic cell. Thesephosphatases dephosphorylate AMPA and NMDA receptors, thus promoting receptor removal from the membrane. (C) During high-frequencystimulation (HFS; often used for LTP induction) strong presynaptic depolarization triggers robust glutamate release promoting Ca2+ influx via AMPAand NMDA receptors into the postsynaptic cell. This Ca2+ influx activates protein kinases (FYN, PKA, PKC, PKMz, and CaMKII) that phosphorylatereceptors and promote stabilization of receptors at the membrane. These kinases also activate local protein synthesis, which leads to insertion of newreceptors in the membrane. (D) By inducing robust glutamate release with HFS up to 300 Hz, activation of postsynaptic group I mGlu receptors leadsto AMPA receptor internalization. (E) The LTP that is dependent on retrograde endocannabinoid (eCB) signaling consists postsynaptic activation thatinduces eCB synthesis, especially increasing 2-AG, which then acts presynaptically on CB1 receptors limiting transmitter release (for more details, seeFig. 12). This form of plasticity can be experimentally evoked, e.g., in the striatal neurons by 100-Hz HFS.


that last for hours to weeks. In contrast to short-termplasticity, the nature of long-term forms of plasticityinvolves both pre- and postsynaptic alterations (Fig. 3).1. Presynaptic Forms of Long-term Plasticity.

Presynaptic LTP has been best studied at hippocampalCA3mossy fiber synapses (Weisskopf et al., 1994; Nicolland Schmitz, 2005), but similar forms of LTP have beenfound inmultiple brain areas (Salin et al., 1996; Castro-Alamancos and Calcagnotto, 1999; Lopez de Armentiaand Sah, 2007). This presynaptic LTP appears at bothexcitatory and inhibitory synapses and does not requirepostsynaptic NMDARs (but see Yeckel et al., 1999).Instead, presynaptic LTP appears to be induced by anactivity-dependent rise of presynaptic residual calciumthat results in a rise in cAMPand subsequent activationof protein kinase A (PKA). This, in turn, modifies thefunctions of proteins that act to coordinate synapticvesicle interactions with the presynaptic active zone,leading to a long-lasting increase in neurotransmitterrelease (Castillo, 2012).Endocannabinoid-mediated long-term depression

(eCB-LTD) is a widely expressed form of long-termplasticity at both excitatory and inhibitory synapses.Brief robust neuronal stimulation triggers the synthesisof eCBs, lipophilic molecules that travel retrogradelyacross the synapse to activate the presynaptic CB1cannabinoid receptors (see section II.H), which sup-presses neurotransmitter release via a wide range ofeffector molecules, including voltage-dependent cal-cium channels (VGCC), potassium channels, PKA, p38mitogen-activated protein kinase (MAPK), and c-JunN-terminal kinases (JNK) (reviewed in Howlett et al.,2002). Importantly, CB1 receptors activation per se isnot sufficient for eCB-LTD induction. Rather, the pre-synaptic terminal integrates multiple signals to gener-ate eCB-LTD (Heifets andCastillo, 2009). CB1 receptorsalso mediate short-term plasticity of DSI and DSE (Fig.2). Interestingly, some hippocampal inhibitory synap-ses can undergo both short- and long-term forms of eCB-mediated plasticity, where the time frame of depression(short-term versus long-term manner) depends on thedownstream signaling pathways (Heifets and Castillo,2009). Particularly, cAMP/PKA-dependent signalinghas been shown to be necessary only for eCB-LTD butnot for DSI (Chevaleyre and Castillo, 2003).2. Postsynaptic Forms of Long-term Plasticity.

NMDAR-dependent LTP and LTD are the best un-derstood forms of long-lasting synaptic plasticity (Blissand Lomo, 1973; Hayashi et al., 2000). Early phases ofLTP and LTD are mediated by a redistribution of botha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acidreceptors (AMPAR) and NMDARs at the postsynapticmembrane and/or by changes in presynaptic transmit-ter release.With time, such changes are consolidated bystructural alterations, which require synthesis of newproteins (Kasai et al., 2010).

In in vitro experiments on acute brain slices, LTP andLTD can be induced by distinct patterns of activity(reviewed in Holscher, 1999; Luscher and Malenka,2012) (Fig. 3). High-frequency stimulation (HFS) of thepresynaptic cell (tetanic pulses at 50100 Hz) is com-monly used to induce LTP. This stimulation protocolcauses a strong postsynaptic depolarization that removesthe Mg2+ block of NMDARs, allowing timed Ca2+ influx.This triggers the downstream molecular cascades in-ducing LTP. In contrast, low-frequency stimulation (LFS,at 0.15 Hz) often induces LTD. Typically, it causes onlya weak postsynaptic depolarization that results in amod-est but prolonged Ca2+ influx triggering the downstreammolecular cascades driving to LTD.Since the original discovery of LTP in the hippocam-

pus (HC) (Bliss and Lomo, 1973), LTP and LTD havebeen observed in a variety of other brain regionsincluding the ventral tegmental area (VTA), NAc,PFC, and amygdala ex vivo (Luscher and Malenka,2011) and in vivo (Canals et al., 2009; Zhang et al.,2015b). Different brain regions appear to exhibit differ-ent forms of LTP and LTD, and therefore, synapsesrecruit different signaling pathways to accomplish theirfunctions. Themost studied forms of neuroplasticity areNMDAR-dependent LTP and LTD, with the NMDARsproviding the major pathway for Ca2+ influx (Huangand Kandel, 1996). There is also NMDAR-independentLTP and LTD, in which VGCC (Kato et al., 2009) orGluA2 subunit-lacking AMPARs (Lamsa et al., 2007)provide the Ca2+ influx triggering induction.In the early phases of LTP, elevated Ca2+ triggers

persistent activation of protein kinases including PKA,Ca2+/calmodulin kinase II (CaMKII), and protein ki-nase C (PKC). A striking feature of CaMKIIa is itscapacity for autophosphorylation at threonine residueThr286, which keeps this kinase activated even in theabsence of Ca2+ (Giese et al., 1998). During this stage,atypical protein kinase C (PKMz) may also becomeautonomously active (Ling et al., 2006; Sacktor, 2008).Autonomously active and other protein kinases usephosphorylation to carry out the twomajor mechanismsunderlying the expression of LTP: first, they phosphor-ylate existing AMPARs and NMDARs to increase theiractivity, and second, they mediate the insertion of newreceptors into the postsynapticmembrane (see below formore detailed description).3. a-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic

acid and N-methyl-D-aspartate Receptor Phosphorylation.Huganir and coworkers made detailed experimentson the regulation of AMPAR trafficking and functionby phosphorylation during LTP/LTD (reviewed inShepherd and Huganir, 2007). During LTP, PKA andCaMKII are recruited to phosphorylate serine residuesin the GluA1 subunit at Ser831 (Mammen et al., 1997)and Ser845 (Roche et al., 1996) and the GluA2 subunitat Ser880 (Chung et al., 2000), promoting receptorinsertion and synaptic potentiation (Malenka and Bear,

880 Korpi et al.

2004; Huganir and Nicoll, 2013). Inhibition of PKA andCaMKII or removal of the above-mentioned phosphor-ylation sites from AMPAR subunits can impair LTP(Lee et al., 2003;Malenka andBear, 2004). Importantly,knock-in mutant mice that lack both Ser831 and Ser845phosphorylation sites on GluA1 subunits demonstratedimpaired memory in behavioral tests (Olivito et al.,2014).Conversely, during LTD, phosphatases calcineurin

and protein phosphatases 1 and 2 are recruited todephosphorylate AMPARs, promoting their removalfrom the synapse (Malenka and Bear, 2004; Huganirand Nicoll, 2013).Phosphorylation of NMDARs is also essential during

LTP/LTD expression because it contributes to stabili-zation of the receptors at the synapse by forming morestable binding to postsynaptic density (PSD) proteins.In fact, Fyn and Src kinases phosphorylate GluN2Bsubunit at Tyr1472, which prevents the clathrin adap-tor protein AP-2 from binding to the GluN2B subunit,thus blocking endocytosis (Zhang et al., 2008).4. a-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic

acid and N-methyl-D-aspartate Receptor Trafficking.During LTP, new GluA2 subunit-lacking AMPARs arerapidly transported to the cell surface under the influ-ence of protein kinases (Malinow and Malenka, 2002;Luscher and Malenka, 2012). Importantly, these newAMPARs exhibit some unique properties: first, theyhave greater single-channel conductance and, second,greater permeability to Ca2+ ions, which facilitatesCa2+-dependent signaling events. Noteworthy, thisearly insertion of new Ca2+-permeable non-GluA2-containing AMPARs to the synapse is independent ofprotein synthesis. This is achieved by having a non-synaptic pool of AMPARs adjacent to the postsynapticmembrane (Kauer and Malenka, 2007). Moreover,NMDARs are also actively replaced during LTP/LTD(Barria and Malinow, 2002). However, the data onNMDARs are still puzzling due to wide variability inexpression, kinetics, and regulation sites (reviewed inShipton and Paulsen, 2014). Particularly interesting forLTP expression are the GluN2B subunit-containingNMDARs because of their special association with CaM-KII. This interaction maintains CaMKII close to itssubstrates, such as AMPARs, to initiate the phosphory-lation events that support synaptic strengthening.Postsynaptic LTD is associated with receptor with-

drawal from the synapses (Hayashi et al., 2000; Belloneand Luscher, 2006). LTD can also be generated byactivation of VGCCs (Christie et al., 1997) or metabo-tropic glutamate receptors (mGlu) (Bellone and Luscher,2006) and does not require NMDAR activation. Althoughthe above cellular model of LTP/LTD presented above isexplained entirely by its postsynaptic mechanisms, addi-tional eventsmay also occur presynaptically, for example,via alteration in probability of glutamate vesicle release

by different retrograde signaling mechanisms (reviewedin Luscher and Malenka, 2012).With time, the insertion/removal of additional recep-

tors is likely to result in the rearrangement of the PSDwith different scaffolding proteins, spine ultrastruc-ture, or even spine density that requires gene transcrip-tion and protein synthesis. In fact, spines associatedwith synapses that underwent LTP were found tobe enlarged (Matsuzaki et al., 2004; Holtmaat andSvoboda, 2009; Kasai et al., 2010), whereas spines thatunderwent LTD were shrunken or even disappeared(Nagerl et al., 2004; Kasai et al., 2010). Although thesestructural changes are not absolutely required forneuroplasticity, the stabilization and maintenance ofexisting synapses needs activity of several kinases suchas PKA, CaMKIV, PKMz, and ERK that induce proteinsynthesis either locally in the dendrites from prefabri-cated mRNAs or by nuclear transcription (Sacktor,2008). These mechanisms are also involved in synaptictagging, an activity-driven molecular labeling of synap-ses to be strengthened (Frey andMorris, 1997; Sajikumarand Frey, 2004; Sajikumar and Korte, 2011).

D. Forms of Long-term Plasticity atInhibitory Synapses

Several forms of NMDA receptor-dependent neuro-plasticity (both LTP and LTD) have been described atinhibitory GABAergic synapses (reviewed in Castilloet al., 2011). Depending on interneuron subtype andbrain region, inhibitory synaptic plasticity results inchanges either in presynaptic GABA release or inpostsynaptic g-aminobutyric acid A (GABAA) receptorresponsiveness. Interneurons also show glutamatergicsynapse plasticity, which has characteristics that arepartly different from those of principal neurons(reviewed in Kullmann and Lamsa, 2011).Presynaptic inhibitory plasticity is mediated via retro-

grade messengers produced in an activity-dependentmanner (by afferent stimulation of excitatory inputs) inthe postsynaptic neuron and transferred back across thesynapse to modulate presynaptic GABA release. Themoststudied presynaptic form of the plasticity is the eCB-mediated inhibitory LTD, in which released eCBs inhibitpresynaptic GABA release by acting on CB1 cannabinoidreceptors (Chevaleyre et al., 2006; Adermark et al., 2009)(Fig. 3). Another form of inhibitory plasticity, NO-mediated inhibitory LTP, requires NO as a retrogrademessenger to stimulate presynaptic guanylyl cyclase,resulting in the potentiation of presynaptic GABA release(Nugent and Kauer, 2008). Brain-derived neurotrophicfactor (BDNF)-tropomyosin receptor kinase B (TrkB)-mediated inhibitory LTP recruits BDNF that activatesits presynaptic receptor TrkB and potentiates GABArelease (Xu et al., 2010). It is evident that GABAergicsynapses from interneuron populations are very plasticand actively participate in brain circuit refinement, learn-ing, and memory formation (Chen et al., 2015).


Postsynaptic inhibitory plasticity is also mediated bydiverse mechanisms in different synapses. Similar toglutamatergic receptors, phosphorylation of postsynap-tic GABAA receptors by protein kinases, including PKA,PKC, CaMKII, and Src, leads to changes in the integralanion channel function (reviewed in Nakamura et al.,2015). The effects of phosphorylation are dynamic(increases and decreases in receptor activity) and de-pendent on the subunit composition of the GABAAreceptor in a fashion that is not fully known at present.Moreover, phosphorylation of cation-chloride cotrans-porters such as Na+-Cl2 cotransporters, Na+-K+-2Cl2

cotransporters, and K+-Cl2 cotransporters (KCCs)affects their activity and consequently the amplitudeof GABAA receptor-mediated responses by altering thetransmembrane anion gradients (reviewed in Kailaet al., 2014). The function of GABAA receptors is furtherdynamically regulated by constitutive exo- and endocy-tosis of receptors at GABAergic synapses (reviewed inMichels and Moss, 2007).

E. Developmental Maturation of the Brain in Rodentsand Humans

Brain development is a delicate and complicated processthat can be adversely affected by drugs during pregnancy,leading to teratogenic effects (reviewed in Manent et al.,2008; Jutras-Aswad et al., 2009; Holbrook and Rayburn,2014; Ross et al., 2015). In the present review, therewill beonly short notes on drug actions on fetal brain develop-ment, such as fetal alcohol effects during pregnancy orearly postnatal period in rodents (Valenzuela et al., 2012).A thorough comparative analysis of early brain matura-tion in different species, including rodents and human,recently appeared (Workman et al., 2013). It is possiblethat vulnerability for drug abuse may be greater theyounger the exposure/use of drugs is started.Developmental neuroplasticity is based on genetic

programs in addition to activity-dependent processes,and after the maturation, the neuroplasticity is expectedto be more homeostatic, dependent on molecular inter-actions in brain cells, neurites, spines, and synapseswithin a relatively rigid extracellular milieu. The rele-vant question to the review of drug-induced neuroplas-ticity is the vulnerability of the brain to drug effectsduring periods of childhood, adolescence, and adulthood.These periods are often defined in humans as adoles-cence, starting at 10 to 12 years of age and ending at 18 to20 years. Childhood precedes it and adulthood comesafter it, and these periods can be further subdivided intoshorter periods by using behavioral and brainmaturationindexes (Bossong and Niesink, 2010; Pressler and Auvin,2013; Semple et al., 2013). For the present review animportant aspect to consider is how rodent models atspecific postnatal ages compare with human childhood,adolescence, and adulthood. Figure 4 gives a roughestimation of how to relate differently timed rodentexperiments to human brain developmental stages. The

figure depicts various brain cell developmental processes,emergence of selective behavioral phenotypes, and thedevelopment of the behavioral modulation of brain pro-cesses by external stimuli in humans and rodents.Several precautions have to be kept in mind, though.

First, there is a large variation between the develop-mental program timescale for different brain regions,with the cortical regions and the cerebellum developinglast (Huttenlocher, 1979; 1990; Huttenlocher andDabholkar, 1997). Second, there are genetic, sex, anddisease vulnerability differences between the phases ofbrain development (Thompson et al., 2001; Vasileiadiset al., 2009; Asato et al., 2010; Brenhouse and Andersen,2011; Dennis and Thompson, 2013). Third, variousneurotransmitter systems develop at different timelines

Fig. 4. Phases of brain development and comparable ages in humans androdents (modified from Semple et al., 2013 and Flurkey and Currer, 2004).(A) Number of spines (per 50 mm of the apical dendrite) of the humanprefrontal cortical pyramidal neurons in layers IIIc and V are shown bycurves estimated from (Petanjek et al., 2011). The figure illustrates normalbehavioral features (in boxes) and the onsets of attention deficithyperactivity disorder (ADHD) and schizophrenia. It is important to notethat puberty in female rats takes place roughly from P28 (vaginal opening)to the first estrus cycle on P40 and in male rats from P40 (balanopreputialseparation) to spermatozoa maturation on P60, respectively (Schneider,2008). (B) Postnatal development of the number of neurons (yellow traces)and nonneuronal cells (blue) for whole rat brain (WB), cerebral cortex (Ctx,with the scale in parentheses), and cerebellum (Cb) (Bandeira et al., 2009;Mortera and Herculano-Houzel, 2012) indicates rapid changes in rodentsduring the first 3 postnatal weeks, the cell numbers remaining ratherstable after that. (C) Timelines for the brain developmental processesimportant for the functions of neuronal and various glial cells.

882 Korpi et al.

in different brain regions, with the DAergic andGABAergic systems of the cortex developing late (Tsenget al., 2007; Hashimoto et al., 2009; Kilb, 2012; Manittet al., 2013;Ouellet anddeVillers-Sidani, 2014). Fourth,various neurotrophic factors and their receptors havedifferential expression profiles in various brain regions(Perovic et al., 2013). Thus, the prolonged effects ofdrugs of abuse are also dependent on the developmentaltime of the exposure to drugs (Stanwood and Levitt,2004). Furthermore, activating and stress-relievingeffects including neonatal handling, maternal care,juvenile play, enriched environment, social contacts orlack thereof, and lesioning effects on the brain also havelong-term effects and their critical times (Meaney et al.,1988; Liu et al., 1997; Caldji et al., 1998; Bouwmeesteret al., 2002; Sale et al., 2007; Tseng et al., 2009; Foscarinet al., 2012). A good example is the differential activa-tion of the caudate-putamen, NAc, and PFC of rats atdifferent ages by a stress-inducing dose of the GABAAreceptor benzodiazepine-site inverse agonist FG-7142(Lyss et al., 1999) known to induce a stress-like activationof the mesocortical DA system in adult rats. The DAsystem has a delayed maturation in the cortex, becomingmature by postnatal day 60 (Kalsbeek et al., 1988). Atpostnatal day 10, only the NAc is strongly activated byFG-7142, at day 18 all regions start to be activated, andfinally at 45 and 100 days of age, the PFC is the brainregion that is activated the most.Potential for neuroplasticity is greater in young than

adult rodents (Hensch, 2005; Bavelier et al., 2010;Hensch and Bilimoria, 2012), as is demonstrated, forexample, by easier induction of the NMDAR-dependentLTP using tetanic stimulation in the NAc neurons of3-week-old than adult mice (Schramm et al., 2002).Furthermore, the brain growth-promoting and limitingresponses to injury are blunted and dysregulated inelderly (2224 months) versus adult (24 months) rats(Li and Carmichael, 2006; Li et al., 2010b).Perhaps an extreme illustration of the prolonged

neurodevelopmental process and normal potential forneuroplasticity in humans may be the profiles of layerIIIc and V spine densities in normal PFC (Petanjeket al., 2011), which is illustrated in Fig. 4A. The spinedensities were determined using Golgi staining of post-mortem brain sections at various ages. The peak densityof dendritic spines is seen at the age of 512 years, afterwhich the density slowly reduces to stable levels after30 years of age. This extended spine-pruning period isconsistent with the results of human functional mag-netic resonance imaging and connectivity studies thathave indicated prolonged developmental changes in thecortical areas (Thompson et al., 2005; Knickmeyer et al.,2008; Dosenbach et al., 2010). A corresponding delayeddevelopment of the structural connectivity between baso-lateral amygdala and mPFC has been observed in the rat(Cunningham et al., 2002), with a wide range of differentexperiences, including the effects of various drugs of abuse,

knownto induce long-lastingplasticity in thePFCalso in theadulthood (Kolb and Gibb, 2015).

F. Developmental Neuroplasticity: Critical Periods,Reopening of Plasticity in Adults

The development and critical period plasticity of thesensory systems, particularly the visual cortex, have beenwell described. In rats and mice, critical periods ofplasticity in the visual, somatosensory barrel, andauditorycortices take place early in the development of thesesensory systems. Importantly, some of the associatedplasticity mechanisms are reused or reopened in adultneuronal plasticity. Drugs inhibiting or stimulating theGABAA receptors can promote or delay the critical periods,respectively (Hensch, 2005). Synaptic pruning eliminatesextra nonfunctional connections formed during initialoverproduction of synaptic connections, while strengthen-ing functionally important ones (Changeux and Danchin,1976). In the normal adult brain, neural plasticity isneeded for maintaining neuronal network excitability,large-scale regulation of cortical and subcortical circuits,and fine-scale experience-dependent refinement andmaintenance of local circuits (Griffen et al., 2012).The extracellular matrix (ECM) network is an active

participant in brain function and neuroplasticity(reviewed in Pavlov et al., 2004). The ECM consists ofvarious molecules (glycoproteins and proteoglycans)secreted by all cells that are assembled inside thismatrix (Bosman and Stamenkovic, 2003). ECMnetworkregulates synaptogenesis, consolidation, strengthening,and maintenance of synapses (Fields and Itoh, 1996).During development, ECM participates in neuronaldifferentiation, neuronal movement, guidance for grow-ing axons, and synaptogenesis (Pavlov et al., 2004). Inthe mature brain, ECM is crucial not only for anchoringof neurons and organization of brain regions (structuralfunction) but also for transducing a wide range ofsignals to the neurons (Thalhammer and Cingolani,2014). Interaction between the ECM and cell is facili-tated by cell adhesion molecules (CAMs), particularlyby integrins. Integrins are heterodimeric transmem-brane glycoprotein receptors that mediate ECM-cellinteractions via binding ECM proteins and cellulartransmembrane proteins (e.g., ion channels and growthfactor receptors). ECM plays an important role inregulating actin polymerization, spine morphology,and GluA2 subunit trafficking in culture (Wu andReddy, 2012). In regard to this review, suppression ofthe synthesis of integrin-linked kinase increases thelevel of Ser845 phosphorylated GluA1 subunit-containing receptors and spine density in the NAc andblocks cocaine-induced sensitization (Chen et al., 2010).Moreover, disruption of integrin expression in the NAcinterfered with trafficking of GluA2 subunit-containingreceptors, which affected cocaine-induced conditionedplace preference (CPP) and reinstated drug seeking(Wiggins et al., 2011).


Perineuronal nets (PPN), first described by CamilloGolgi (reviewed in Celio et al., 1998), develop aroundspecific neuronal populations to protect highly activeneurons and to preserve their structure in anexperience-dependent manner during pruning of extraconnections in the visual cortex and many other parts ofthe central nervous system (CNS) (Kwok et al., 2011;Soleman et al., 2013; Ye and Miao, 2013). PNNs areprimarily composed of chondroitin sulfate proteogly-cans (e.g., aggrecan, versican, neurocan, brevican, andphosphacan), hyaluronan, link proteins such as cartilagelink protein 1 Crtl1 or Hapln1 (Carulli et al., 2010), andtenascins. In the cortex, PNNs selectively surround theparvalbumin (Pv)-containing fast-spiking GABAergicinterneurons (Ye and Miao, 2013), starting after post-natal day 10 (P10) and achieving adult levels by P42.Thus, the PNNs develop during the critical period, andafter closing, the critical period for plasticity can bereopened by degradation of the PNNs (Pizzorusso et al.,2002). Although the negatively charged components of thePNN can have multiple molecular interactions and func-tions, for the maturation of Pv-interneurons and regula-tion of critical periods, the PNN provides a high-affinitybinding site for the lysine-arginine containing domain ofthe transcription factor Otx2 (orthodenticle homeobox 2)(Beurdeley et al., 2012). Once Otx2 reaches the Pv-interneurons of the visual cortex, being synthesized inthe retina or the lateral geniculate of the visual tract(Sugiyamaet al., 2008), it induces expression ofPv,GAD65(glutamate decarboxylase 65 kDa), GABAA a1 subunit,and Kv3.1b K

+ channels specifically in these targetneurons (Sugiyama et al., 2008). This process is experiencedependent, that is it cannot begin before eye opening, afterwhich the critical period for setting ocular dominance andvisual acuity is started (Fagiolini and Hensch, 2000;Sugiyama et al., 2009). Pv-interneurons in the visualcortex mature in connections (e.g., forming perisomaticsynapses on principal neurons) and intrinsic firing prop-erties (e.g., fast spiking) by the end of the critical period,thus allowing the firm functional cortical connections forbinocular visual acuity to be formed (Kuhlman et al.,2010). PNNs interact with receptors such as NgR1 (NoGoreceptor) and leukocyte common antigene-related phos-phatase (Akbik et al., 2013; Ye and Miao, 2013).PNN formation in the basolateral amygdala (BLA)

stabilizes fearmemories in adults, unlike in thedevelopingbrain when they can be readily extinguished (Gogollaet al., 2009). Degradation of PNN by the enzyme chon-droitinase ABC in the BLA, left the fear conditioningintact, but after extinction the animals failed to spontane-ously recover the fear and to express freezing under thefear-related context, indicating erasure of the memory bydegradation of the PNN. Two recent papers indicate thatPNNs affect drug- and experience-induced memory inrodents. First,Karpova et al. (2011) showed that fluoxetinetreatment during extinction training in fear-conditionedadult mice led to erasure of fear memory, associated with

a reduction in the proportion of Pv-containing interneu-ronswithPNNexpression in theBLA,HC, and infralimbiccortex (IL). In the BLA, fluoxetine treatment also in-creased the expression of polysialynated neuronal celladhesion molecule (PSA-NCAM) and decreased KCC2levels in the BLA and HC, with all these data suggestingthat fluoxetine treatment with extinction training induceda juvenile-like state in the interneurons of the fearcircuitry. Second, infusions of chondroitinase ABC intothe BLA or the central nucleus of the amygdala (CeA) ofadult rats during extinction of morphine- and cocaine-induced CPP prevented priming-induced reinstatement(Xue et al., 2014). This effectwas associatedwith increasedlevels of AMPAR GluA1 and GluA2, but not GluA3, andincreasedBDNF levels in theBLA. Thus, the extracellularmatrix and its restructuring seem to be involved in themodulation of specific memories that can also be associ-ated with persistent drug effects. Perhaps, changes in thematrix may be developed into future therapies for enhanc-ing or reversing plasticity (Castren et al., 2012). However,we should learn more about how the extracellular matrixinduces or restricts plasticity and when would be the righttime to increase or decrease the matrix components inappropriate neuronal circuits.One of the major anionic extracellular components

around mature interneurons is formed by PSA-NCAMmolecules with antiadhesive properties. Expression ofPSA-NCAM has been described in a subpopulation ofGABAergic calbindin-positive and somatostatin-positive neurons in the adult rat and mouse cerebralcortex, without expression in the principal neurons orother interneuron types (Gomez-Climent et al., 2011).These PSA-NCAM-positive interneurons had reducedsynaptic connections as judged from confocal immuno-histochemical counting of synaptophysin-positive peri-somatic and peridendritic puncta in close apposition toGAD67-expressing cells. This suggests that expressionof PSA restricts the synaptic connections of the inter-neurons in mature brain. Interestingly, DAergic ago-nists increase (Castillo-Gomez et al., 2008) and alsoserotonergic drugs increase (or downregulate stress-elevated) (Wedzony et al., 2013) production of PSA inthe mPFC. NCAM is needed for hippocampal LTP thatis deficient in NCAM-KO mice (Caroni et al., 2012;Stoenica et al., 2006). Intraventral mPFC infusion ofendoneuraminidase, which cleaves PSA from NCAM tomice that were trained to extinguish operant alcohol-seeking behavior, strongly attenuated extinction (Barkeret al., 2012), suggesting that reduced levels of PSA-NCAM restricts the plasticity needed for extinction.Similarly to PNNs, extracellular NCAM and PSA-NCAM mechanisms also participate in retaining themature structures, although they are also needed fornovel connections and pathways of plasticity. They dooffer a number of possible, but still deficiently character-ized, targets for drug action, and they may be importantfor understanding impairments of interneuron function

884 Korpi et al.

in depression and schizophrenia (Nacher et al., 2013;Wedzony et al., 2013).

G. Neurotrophins as Regulators of Neuroplasticity:Brain-Derived Neurotrophic Factor as an Example

BDNF is an important trophic factor for nerve cells inthe CNS (Park and Poo, 2013), and it serves here as anexample of themultiple roles that the neurotrophins areplaying in orchestrating development, maintenance,and plasticity of neurogenesis and neural connections.It is secreted from various cells activity dependentlyand exerts effects on local synaptic plasticity.In short, mature BDNF is a 14-kDa protein that is

cleaved from prepro-BDNF expressed both in neuronsand glial cells in the CNS. BDNF gene has IX promotersin humans and rodents, and at least the promoter IV isactivity dependent with Ca2+ and cAMP responseelements (Park and Poo, 2013). BDNF preferentiallyacts via activation of tropomyosin receptor kinase B(TrkB) (Minichiello, 2009). Like many of the trophicfactor receptors, TrkB belongs to the family of receptortyrosine kinases. Signaling of BDNF via TrkB leads toactivation of Ras-ERK, phosphatidyl inositol 3-kinase(PI3K) and phospholipase C, gamma (PLCg) pathways,which serve for survival, differentiation, and synapticplasticity phenomena in neurons (Minichiello, 2009).Pro-BDNF activates preferentially P75 neurotrophinreceptor and often has opposite or different effects thanmature BDNF (Lu et al., 2005).BDNF has important interactions with the GABA

system. It is needed for the GABA system to becomeinhibitory during development (Huang et al., 1999), butits activation in the mature nervous system may lead toreduction of KCC2 expression and subsequent change toGABAergic excitation. This has been described duringdevelopment of neuropathic pain in the spinal cord, withactivated microglial cells releasing BDNF, which makesthe circuitry abnormally sensitive to stimuli due todownregulation of KCC2 expression (Coull et al., 2005)and due to LTP induction via GluN2B upregulation(Ding et al., 2014). In similar preclinical models, certainbenzodiazepine-site agonists, the efficacy of which issupposed to be dependent on KCC2-driven Cl2 gra-dients, surprisingly show strong analgesic effects byablating neuropathic pain symptoms (Knabl et al.,2008; Paul et al., 2014a), stressing that multiple parallelpathways are involved (Zeilhofer et al., 2012). The KCC2downregulation andGABAergic excitation also appear toswitch rewarding responses in the CPP test to intra-VTAGABAA agonist muscimol from DA-sensitive to DA-insensitive (or vice versa for the antagonist bicuculline)in opioid-dependent rats (Laviolette et al., 2004). Thiscondition could be mimicked by intra-VTA infusion ofBDNF (Vargas-Perez et al., 2009). Furthermore, morerecently it was shown that in stressed animals, specifichypothalamic neurons are excited byGABAdue to stress-induced phosphorylation and endocytosis of KCC2

(Sarkar et al., 2011). These striking results appear tolink chronic drug exposure, stress, and aversive painstimuli to pathologic neuroadaptation in BDNF andGABA systems and indicate that boosting the chloridegradients might be an attractive drug developmenttarget (Gagnon et al., 2013).Although GABAA receptor-mediated excitation or

depolarization in principal neurons is known to beimportant in early neuronal development (Ben-Ariet al., 2007; Blaesse et al., 2009) and perhaps in specificpathologic conditions in the mature nervous system(Medina et al., 2014), to our knowledge there is stilllittle evidence that drugs acting on GABAA receptorswould induce tolerance via reversed Cl2 gradientsduring in vivo treatments. In addition, acute behavioraleffects of drugs with preferential action on extrasynap-tic GABAA receptors, including gaboxadol, were notaffected in mice hypomorphic to KCC2, although theresponses to the synaptic GABAA receptor agonistdiazepam were strongly blunted (Tornberg et al.,2007). Interneurons express KCC2 already early indevelopment (Batista-Brito et al., 2008), consistent withshunting inhibition in the interneurons already atpostnatal days ,10 (Banke and McBain, 2006). Thestabilized role of specific interneuron populations, suchas Pv-interneurons, may become even more importantat the end of the critical periods (Hensch, 2014).7,8-Dihydroxyflavone, a specific small molecule ago-

nist of TrkB receptors, has a wide activity in variouspreclinical tests for cognitive and emotional behaviors(Baker-Andresen et al., 2013), but to our knowledge, ithas not yet been tested for modulation of drug-inducedplasticity. Perhaps only together with a systemicallyactive antagonist of TrkB receptors, the roles of BDNF-TrkB signaling in various phases of drug-inducedplasticity would become better established. However,as mentioned above, in the developing CNS during thecritical periods of cortical circuitry formation BDNFsignaling is indispensable for maturation of GABAergicinhibition (Huang et al., 1999) (reviewed in Kuczewskiet al., 2010).As discussed above, inhibitory interneurons have an

important role in neuroplasticity, and their diversity,development, and functions were recently characterizedin the cortical regions (Klausberger and Somogyi, 2008;Ascoli et al., 2008; Lapray et al., 2012), but only initiallyin the midbrain (Olson and Nestler, 2007; Nair-Robertset al., 2008). Because the midbrain GABAergic inter-neurons even have different embryonic origins to thecortical interneurons, likely subpopulations and char-acteristics of midbrain interneurons will be an interest-ing target for research and drug effects.

H. Animal Models of Drug Reinforcementand Addiction

Animal models are critical for understanding theneurobiological basis of drug actions and drug addiction.


Because addiction is a human phenomenon, there areno complete animal models for it. However, somecharacteristics of this syndrome can be satisfactorilymodeled in laboratory animals. It is generally agreedthat when viewed as experimental preparations forstudying the human syndrome, animal models exhibitgood construct and predictive validity, even if their facevalidity can be disputed (for a comprehensive review,see Sanchis-Segura and Spanagel, 2006).Conceptualizations of the development of addictionhold

that drugs are initially taken voluntarily because of theirpositive reinforcing or rewarding effects, and thereforemodels have been constructed for evaluating the ability ofdrugs to act as reinforcers or conditioned reinforcers,including a variety of animal models of drug SA. Behav-iorally, they can be classified as operant models based onoperant conditioning and nonoperant models that arerestricted to various oral SA procedures used particularlyfor measuring ethanol consumption. Traditionally, thesemodels were based on a two-bottle choice between waterand ethanol solution, but it was subsequently found thatmanipulation of the temporal access to ethanol canincrease the consumed ethanol during drinking bouts,leading to development of binge-drinking models (Wise,1973; Rhodes et al., 2005; Simms et al., 2008).Operantmodels arebased on thedeliveryofa reinforcer

contingently to the completion of the required responsethat is determined by the reinforcement schedule. Re-inforcement schedules exert powerful control over behav-ior and can be used for evaluating the nonspecific andmotivational drug effects. For example, progressive ratioschedules can measure the reinforcing efficacy of a drug,whereas the second-order schedules are useful for study-ing the role of conditioned reinforcers in maintainingbehavior. Various routes of SA can be used, e.g., in-travenous, intracranial, intragastric, or oral. In particu-lar, the intravenous drug SA animal model is consideredto be predictive of drug abuse potential in humans.In conditioned preference paradigms, drugs effects (un-

conditioned stimulus) are paired repeatedly with a pre-vious neutral stimulus. Through Pavlovian conditioning,the neutral stimulus acquires the ability to act as a condi-tioned stimulus.Most commonly, conditioned preference isstudied using CPP procedures. These procedures use anapparatus with two or more separate distinctive compart-ments. Access to one compartment is paired with druginjections, whereas the other compartment is accessibleafter vehicle injections. After repeated pairings of the drugand the vehicle with their compartments, the animal isallowed tomove freely in all compartments during the testsession. Increase in the time spent in the drug-associatedcompartment is taken as a measure of CPP.Animals will perform tasks to self-administer elec-

trical trains of stimulation to many different brainareas, particularly along the medial forebrain bundle,leading to the idea that direct activation of brain revealsthe circuitry involved in reinforcement from natural

rewards. Various intracranial self-stimulation (ICSS)procedures have been used for mapping the neuralcircuitry mediating drug reinforcement and pharmaco-logical suppression thereof. Many abused drugs acutelydecrease the ICSS reward threshold, whereas increasedthresholds have been seen during withdrawal afterinduction of dependence.Compulsive drug seeking and relapse to drug use are

considered to be the hallmarks of addiction, andtherefore attempts have been made to model theseaspects of addiction in animal models. Particularly formodeling drug-seeking behavior, reinstatement proce-dures are now being commonly used. These proceduresconsist of the initial phase of operant drug SA, theextinction phase during which operant responses haveno programmable consequences, and the reinstatementphase during which the extinguished responding isreinstated either with a priming dose of the drug,conditioned stimuli associated with the drug, stressfulfootshocks, or pharmacological stressors. Drug seekinginduced by re-exposure to drug-associated cues hasbeen shown to progressively increase over severalweeks after withdrawal from drug SA, a phenomenonthat has been called incubation of drug craving (Luet al., 2004), which is reminiscent of the alcoholdeprivation effect found earlier in rats after a period ofabstinence (Sinclair, 1972). The compulsive features ofdrug taking, such as resistance to adverse consequencesand escalated intake, are accentuated in models in whichthe daily access to a drug is extended (Ahmed and Koob,1998; Deroche-Gamonet et al., 2004; Vanderschuren andEveritt, 2004).Psychomotor or behavioral sensitization refers to the

progressive increase in locomotion when abused drugsare administered repeatedly. This phenomenon is inessence nonassociative, but the context of drug admin-istration has been demonstrated to have a role in theinduction and expression of sensitization. It has beentheorized that psychomotor sensitization also entailsincreased incentive salience of stimuli associated withdrug effects, and therefore sensitization is implicated intransition to compulsive drug seeking and taking(Robinson and Berridge, 1993).

II. Actions and Persistent Effects of SpecificDrugs of Abuse

A. Cocaine, a Stimulant and Local Anesthetic

Cocaine, derived from the leaves of the coca plant, isa very potent psychostimulant. Cocaine and its deriva-tives, can be inhaled, injected into the bloodstream, orsmoked, with the duration and intensity of effectsdepending on the route of administration. Its effectsinclude euphoria (which might eventually be replacedby anxiety), hyperactiveness, suppression of appetite,local anesthesia, and possible sudden death dueto cardiac arrest that makes it considerably more

886 Korpi et al.

dangerous than other psychostimulants (for compre-hensive review, see Grabowski, 1984). In addition,cocaine has been associated with a variety of cardiovas-cular disorders, including myocardial infarction, heartfailure, cardiomyopathies, arrhythmias, aortic dissec-tion, and endocarditis (Schwartz et al., 2010).Cocaine is a DA, norepinephrine (NE), and serotonin

reuptake inhibitor (Fig. 5; Table 1). It also blocksvoltage-dependent sodium channels, which produceslocal anesthesia, and also has been shown to causelocomotor depression rather than stimulation (Reithet al., 1985). Cocaine increases plasma catecholaminelevels; therefore, it exhibits sympathomimetic proper-ties. As a consequence, it is a potent vasoconstrictor ofblood vessels in the brain (Kaufman et al., 1998; Volkowet al., 1996; Wallace et al., 1996). The addictivepotential of cocaine is widely considered to be mainlybecause of its block of DA reuptake (Sulzer, 2011). Thevast majority of studies have focused on elucidating themechanisms via which DA reuptake suppression leadsto addictive behaviors.In this regard, the holy grail of the field has been to

understand lasting changes induced in the brain due tothis reuptake inhibition, which outlasts the presence ofthe drug in the system. The main rationale for thisapproach is that addictive behaviors persist well beyondthe presence of the drug in the system. One idea thathas gained considerable experimental support in thepast decade or so is that the normal learning processesin the brain (involving the mesocorticolimbic DA sys-tem) are "hijacked" in addiction to reinforce the acqui-sition of the addictive drug. As is discussed in thisreview, experimental results have lent considerablesupport to this notion; however, recently emergingevidence indicates this is only part of the picture (Nuttet al., 2015). Furthermore, emerging evidence indicatesthat other structures play a role in addiction as well. Inthis review, the effects of cocaine on lasting changes inthe brain are discussed in a brain area-dependentfashion, because the evidence indicates different brainregions might play different roles in addiction (Thomaset al., 2001; Ungless et al., 2001; Conrad et al., 2008;Kalivas, 2009; Koob and Volkow, 2010; Luscher andMalenka, 2011; Pascoli et al., 2012, 2014).1. Persistent Ventral Tegmental Area Neuroplasticity

after a Single Dose. The VTA, along with the adjacentsubstantia nigra (SN), forms the principle source ofDAergic projection neurons in the brain. Located in themidbrain, it projects to the many forebrain regionsincluding the ventral striatum (NAc), amygdala, lateralhabenula (LHb), mPFC, ventral hippocampus (vHC),and bed nucleus of stria terminalis (BNST) (see Fig. 1).Although the majority of its projections are DAergic innature, significant populations of GABAergic and glu-tamatergic projection neurons have been observed aswell (Nair-Roberts et al., 2008; Stuber et al., 2010;Hnasko and Edwards, 2012). It is believed that the VTA

plays a role in assigning "salience" to rewarding orpunishing events, and the more salient (or surprising)an event is, the greater the chance that it is learned.DAergic neurons in the VTA drastically increase firingin response to unexpected rewards or punishments andthe cues associated with them (Mirenowicz and Schultz,1996; Ungless et al., 2004; Brischoux et al., 2009).Therefore, a drug that evokes a persistent increase inDA neuronal firing or leads to a long-lasting potentia-tion of excitatory inputs to DA neuronspossibly byinducing LTPcould in theory drive motivationalbehaviors such as are observed toward salient events(Wolf, 2002).Initial evidence indicated that a single, noncontin-

gent (experimenter administered) dose of cocaine led to

Fig. 5. The effect of stimulant drugs on dopaminergic neurotransmissionis characterized by elevation of synaptic dopamine (DA) levels andsubsequent increases in postsynaptic neuronal responses. Cocaine blocksthe dopamine transporters (DAT), resulting in increased synaptic DAlevels due to competitive reuptake inhibition. Amphetamine and relatedsubstances (AMPH), such as methamphetamine, are taken up into thepresynaptic neuron via the DAT, also resulting in increased synaptic DAlevels due to competitive reuptake inhibition. Because AMPHs arelipophilic weak bases they are also capable of entering the cytoplasm viadiffusion through the membrane. Furthermore, once inside the cytoplasmthey enter secretory vesicles and cause an efflux of DA from them.AMPHs are capable of reversing the function of DAT, causing furtherincreases in synaptic DA levels. In contrast, cocaine and relatedstimulants do not enter the presynaptic neuron and enhance synapticDA levels only by inhibiting its reuptake. Either way, the end result isenhanced signaling through DA D1- and D2-like receptors, leading toactivation or inhibition, respectively, of adenylate cyclase (AC) andfurther downstream cellular responses.


a potentiation of excitatory synaptic transmission ontoDA neurons in the VTA, and the changes observed inresponse to repeated cocaine administration were sim-ilar (Ungless et al., 2001; Saal et al., 2003; Borglandet al., 2004). This potentiation was observed by mea-suring the ratio of AMPAR to NMDAR-mediated cur-rents (AMPAR/NMDAR ratio) in a neuron clamped ata fixed voltage ex vivo, 24 hour after cocaine had beenadministered in vivo. One advantage of this protocol,which has been widely used since in the field for cocaineand other drugs of abuse, is that lasting changes inducedin vivo can be observed (as the drug has cleared thesystem). This potentiation of excitatory synaptic trans-mission returns to normal levels when assessed a weeklater (Ungless et al., 2001). This astounding findingindicates that a single dose of cocaine might be sufficientto induce lasting synaptic changes in the VTA andpossibly alter activity in downstream targets. In thisregard, a recent report further characterized the earlierfindings and demonstrated that only VTA DA neuronsprojecting to the NAc shell and not to the mPFC arepotentiated on exposure to cocaine (Lammel et al., 2011).Glutamate receptor neuroplasticity could not be elicitedin the VTA-mPFC neurons even by repeated exposure tococaine, whereas, in contrast to cocaine, an aversivestimulus (a formalin injection) potentiated the synapsesof the mPFC projection in the VTA. Furthermore,optogenetic activation of inputs from the laterodorsal

tegmental nucleus (LDTg, which preferentially synapseonto VTADA neurons projecting to the NAc lateral shell)elicits reward, whereas activation of inputs from the LHb(preferentially synapsing on DAergic projections to themPFC and GABAergic neurons in the VTA) results inaversion (Lammel et al., 2012). These results suggest thatonly the VTA DA neurons involved in reward processingare affected (further discussed within this section con-cerning cocaine effects in the NAc and mPFC).The increase in AMPAR/NMDAR ratio is due to

increased AMPAR currents, which is mainly due toinsertion of Ca2+-permeable GluA2 subunit-lackingAMPAR receptors (Fig. 6) (Mameli et al., 2011). How-ever, NMDAR-dependent currents are reduced as well(reviewed in Luscher, 2013). This reduction is due toinsertion of quasi-Ca2+-impermeable NMDARs contain-ing GluN3A and GluN2B subunits. It has been hypoth-esized that GluN3A-containing NMDARs might playa role in cocaine-evoked AMPAR plasticity as well(Yuan et al., 2013). Therefore, removing the GluN3A-containing NMDARs might reduce cocaine-evokedAMPAR plasticity and mitigate its effects on down-stream targets. Activation of mGlu1 receptors in theVTA removed these NMDARs (Yuan et al., 2013).Furthermore, functionally overriding mGlu1 receptorsin vivo made cocaine-evoked synaptic potentiation in theVTA persistent (that is, it did not return to normal after7 days) (Mameli et al., 2009). This suggests mGlu1

TABLE 1Inhibition of transporter ligand binding and reuptake by cocaine and other stimulants

(A) The concentration ranges of the amphetamines AMPH, METH, and MDMA; the cathinones methcathinone, 4-MMC, and MDMC as well as cocaine and MPH needed todisplace typical DAT, SERT, NET, and VMAT2 ligands (Ki in mM). (B) The concentration ranges required to inhibit monoamine reuptake via the same transporters (IC50 inmM). Note that methodological differences exist between the studies reported in the table.

Drug DAT SERT NET VMAT2 References

Date post:	07-Jan-2016
Category:	Documents
Upload:	jared-feldman
View:	238 times
Download:	1 times

Mechanism of Brain Changes in Addiction

Documents