cocaine and chromatin

8/16/2019 Cocaine and Chromatin

1/3

478 VOLUME 18 | NUMBER 4 | APRIL 2015 NATURE NEUROSCIENCE

N E W S A N D V I E W S

early studies of cerebral metabolism1. A similarsituation is found in patients with SLC2A1 defi-

ciency, a rare genetic disease associated withprominent neurological deficits, in whom a

ketogenic diet is highly beneficial14.Could GLUT1 be used to develop new thera-

peutic interventions in AD? If a reduction inendothelial GLUT1 enhances AD pathology,

it is conceivable that restoring GLUT1 levelscould ameliorate brain dysfunction and damage

in AD. To begin to address this question,

Winkler et al.4 performed adenoviral gene trans-fer in APP Sw mice deficient in Slc2a1. They found

that restoration of GLUT1 in the hippocampusgreatly reduced local Aβ levels. Similar results

were obtained with viral gene transfer of LRP1,the Aβ vascular transport protein suppressed by

GLUT1 deficiency. Although the authors did notdemonstrate rescue of neuronal function and

behavior, the findings provide proof of principlethat upregulation of GLUT1 clears the brain of

amyloid and could have beneficial effects.

Little is known about the mechanism caus-

ing GLUT1 dysregulation in AD. GLUT1expression is controlled by hypoxia-inducible

factor 1 (HIF1α,β). Given that HIF1α is down-regulated in AD15, it is conceivable that HIF1α

suppression leads to reduced GLUT1 expres-sion. However, earlier studies have indicated

thatSLC2A1

mRNA is not reduced in AD,implicating post-translational mechanisms8.

Thus, further studies on the molecular bases ofGLUT1 reduction are needed to provide some

indication of how to counteract it.Irrespective of the many questions outstand-

ing, the data of Winkler et al.4 demonstrate a

multifaceted role of glucose transport in themaintenance of brain structure and function, and

unveil a damaging interaction with AD pathol-ogy. This may open new therapeutic avenues for

this devastating neurodegenerative disease.

COMPETING FINANCIAL INTERESTS

The author declares no competing financial interests.

1. Hoyer, S., Oesterreich, K. & Wagner, O. J. Neurol. 235,143–148 (1988).

2. Harik, S.I., Kalaria, R.N., Andersson, L., Lundahl, P. &Perry, G. J. Neurosci. 10, 3862–3872 (1990).

3. Mamelak, M. J. Alzheimers Dis. 31, 459–474 (2012).4. Winkler, E.A. et al . Nat. Neurosci. 18, 521–530

(2015).5. Nordberg, A., Rinne, J.O., Kadir, A. & Långström, B.

Nat. Rev. Neurol. 6, 78–87 (2010).6. Jagust, W.J. et al. J. Cereb. Blood Flow Metab. 11,

323–330 (1991).7. Kalaria, R.N. & Harik, S. J. Neurochem. 53,

1083–1088 (1989).8. Mooradian, A.D., Chung, H.C. & Shah, G.N.Neurobiol.

Aging 18, 469–474 (1997).9. Jack, C.R. et al. Lancet Neurol. 12, 207–216 (2013).

10. Bateman, R.J. et al. N. Engl. J. Med. 367, 795–804(2012).11. de le Monte, S.M. & Tong, M. Biochem. Pharmacol.

88, 548–559 (2014).12. Krikorian, R. et al. Neurobiol. Aging 33,

425.e19–425.e27 (2012).13. Reger, M.A. et al. Neurobiol. Aging 25, 311–314 (2004).14. Pearson, T.S., Akman, C., Hinton, V.J., Engelstad, K. &

De Vivo, D.C. Curr. Neurol. Neurosci. Rep. 13, 342(2013).

15. Liu, Y., Liu, F., Iqbal, K., G rundke-Iqbal, I. &Gong, C.-X. FEBS Lett. 582, 359–364 (2008).

Endothelial GLUT1deficiency

Cognitivedeficits

Hypoperfusion

Synaptic dysfunction and neurodegeneration

Altered homeostasis Amyloid pathologyEnergy deficit

Microvascularrarefaction

Reduced Aβclearance

Aβ

LRP1

Reduced brainglucose uptake

Glucose

BBB leakage

Figure 1 Mechanisms of brain dysfunction and damage caused by GLUT1 deficiency. Endothelial

GLUT1 deficiency leads to reduced brain glucose transport, vascular rarefaction and disruption of

the BBB, as well as reduced Aβ clearance by suppressing vascular LRP1 expression. These events

result in energy deficit, reduced cerebral blood flow (hypoperfusion), altered homeostasis of the

brain microenvironment and enhanced amyloid pathology. The resulting synaptic dysfunction and

neurodegeneration in turn lead to cognitive deficits.

Anne E. West is in the Department of Neurobiology,

Duke University Medical Center, Durham, North

Carolina, USA.

e-mail:[email protected]

Cocaine shapes chromatin landscapes via Tet1

Anne E West

Chronic cocaine exposure induces long-lasting, transcription-dependent changes in neuronal function. A genome-wide

sequencing study shows how cocaine changes the epigenome to exert specific, long-lasting effects on neuronal transcription.

Memories are the essence of a life. Ask your-

self “Who am I?” and you trigger a mental

movie of schoolrooms, a wedding kiss, a

tasty French pastry or a sled. Neuroscientistshave long sought to understand how past

experiences are encoded in the brain.Recently, the field has fallen in love with

the idea of epigenetics, in which the brainplasticity narrative of dynamic learning

and persistent memory finds a physical

instantiation in the biochemical modifica-tions of genomic DNA and its associated

histone proteins. Methylation of DNA isstrongly associated with persistent biologi-

cal processes in cells, such as X chromosomeinactivation and gene imprinting. When


2/3

NATURE NEUROSCIENCE VOLUME 18 | NUMBER 4 | APRIL 2015 47 9


it was discovered that DNA methylation

can be dynamically lost through the action

of a family of enzymes called the Tets, thissuggested a mechanism by which environ-mental experience could be remembered

via its effect on epigenetic chromatin regu-lation, gene expression and neuronal func-

tion on a behaviorally relevant timescale.Now, in this issue of Nature Neuroscience,

Feng et al.1 provide new evidence for thebiological functions of Tets in brain plastic-

ity, reporting that Tet1 is a target of regula-tion by cocaine and that decreases in Tet1

remodel the chromatin landscape in waysthat change the expression of functionally

important neuronal genes.In mammalian cells, the methylation of

cytosines (5mC) in genomic DNA is mediatedby a small family of DNA methyltransferases.

Cytosine methylation has been predominantlystudied in the context of CpG dinucleotides,

although in the brain non-CpG methylationseems likely to be important as well2. In 2009,Kriaucionis and Heintz3 made the intriguing

discovery that DNA from Purkinje cell nucleicontained an unusual DNA nucleotide, which

they identified as 5-hydroxymethylcytosine(5hmC). In parallel, Tahiliani et al.4 iden-

tified 5hmC in embryonic stem cells and

demonstrated that the human TET1 enzymecatalyzed the conversion of 5mC to 5hmC. Inaddition to Tet1, Tet2 and Tet3 were found to

have similar enzymatic activities, and knock-out and knockdown studies quickly confirmed

the requirement for these enzymes in the gen-eration of 5hmC during cellular differentiation

and embryonic development5–7.The brain contains some of the highest

levels of 5hmC in the body, and they increasefrom early postnatal development through

adulthood, suggesting a role for 5hmC inmature brain function8. Consistent with

this possibility, Tet1 knockout mice show

diminished expression of activity-regulated

genes, abnormal hippocampal long-termdepression and impaired memory extinc-tion, although the connection between these

phenotypes and environmentally drivenchanges in Tet1 function or 5hmC distri-

butions remains unknown9. Furthermore,Tet3 expression is enhanced in infralimbic

cortex following fear conditioning and Tet3knockdown in this brain region is associated

with impaired fear conditioning, furthersuggesting a link between Tets, 5hmC

and transcription-dependent behavioralplasticity 10.

Repeated cocaine exposure can lead toaddiction, which is one of the most persis-

tent forms of environmentally induced andtranscription-dependent brain plasticity.

Thus, Feng et al.1 asked whether cocaine

regulates the Tets in the nucleus accum-bens (NAc), a brain region required for the

rewarding effects of cocaine. Mice that hadbeen repeatedly exposed to cocaine had sig-

nificantly less Tet1 mRNA and protein in theNAc than controls, whereas levels of Tet2 and

Tet3 were unchanged. TET1 mRNA expres-sion was also reduced in the NAc of brains

from humans addicted to cocaine, suggesting

the relevance of this regulatory pathway foraddiction. Knocking down the expression ofTet1 in the NAc enhanced cocaine-induced

conditioned place preference, a behavioralassay of the rewarding effects of cocaine. By

contrast, overexpression of Tet1 in the NAcimpaired conditioned place preference.

These data show that the cocaine-dependentdecrease in Tet1 expression is required for

behavioral plasticity induced by repeatedexposure to cocaine, and, overall, these data

indicate that Tet1 functions to negativelyregulate cocaine reward.

Despite the reduced expression of Tet1in the NAc after cocaine, the authors found

no effect of cocaine on the global levels of

either 5hmC or 5mC as a percentage of totalcytosine in the NAc. However, when they

performed genome-wide sequencing for thedistribution of 5hmC, they found that cocaine

induced significant changes (both increasesand decreases) in 5hmC levels at more than

11,000 sites across the genome, distributedboth in intergenic regions and across gene

bodies. These data are consistent with amodel in which local recruitment of Tet1 to

specific gene regulatory elements mediatesthe regulation of a discrete set of target genes.

To test this model, the authors compared thecocaine-induced changes that they observed

in 5hmC levels at different kinds of genomicelements with alterations in the expression of

nearby genes.In the intergenic regions, 5hmC was

enriched at elements marked by acetyla-tion of histone H3 on Lys27 and monom-

ethylation on Lys4, which are the histonemodifications most strongly associated

with active distal gene enhancers. Eventhough cocaine decreases Tet1 expression

and Tet1 promotes the conversion of 5mCto 5hmC, cocaine exposure was associated

with an equal number of enhancers show-ing increases in 5hmC compared to those

showing decreases in 5hmC. The loss of Tet1was sufficient to mediate increases in 5hmC:

when the authors knocked down Tet1 in the

NAc in the absence of cocaine exposure, theyobserved enhanced 5hmC at several loci thatalso showed cocaine-induced enhancement.

The mechanism by which loss of Tet1 leadsto increased 5hmC was not resolved in this

study, but the authors speculate that it couldarise from secondary dysregulation of 5hmC

metabolism, non-enzymatic contributions ofthe Tets to chromatin regulation11, or func-

tional compensation by Tet2 and/or Tet3.Regardless, as the authors did not detect a

correlation between cocaine-induced 5hmCchanges at enhancers and global changes in

gene expression profiled by RNA-seq, the

functional importance of these chromatinchanges remains unknown.

Over gene bodies, 5hmC was depleted

at transcriptional start sites and enrichedboth upstream of transcription ending sites

and in regions flanking exon boundaries,all of which are consistent with previous

reports8,12. Following cocaine, the authorsfound that changes in 5hmC were enriched

near splice sites. Furthermore, by comparing5hmC distributions at exon boundaries with

RNA-seq data, they found an intriguing cor-relation between changes in 5hmC levels and

Figure 1 Cocaine-dependent changes in the architecture of the 5hmC landscape are associated

with changes in gene expression. In the NAc, 5hmC is enriched at active gene enhancers, across

gene bodies, and near exon boundaries. Enhancers are indicated by peaks of histone H3 modified

by acetylation at lysine 27 (H3K27Ac) and by monomethylation at lysine 4 (H3K4me1). Promoters

are indicated by peaks of histone H3 modified by trimethylation at lysine 4 (H3K4me3) and

transcription starts sites are shown by the arrows. Cocaine exposure reduces Tet1 expression

and results in reduced (shown) or increased 5hmC deposition at both enhancers and splice

sites. Changes in 5hmC at exon boundaries (*) are associated with changes in splice site usage.

Alternative splicing of an mRNA that generates transcript A-B-C-D before cocaine generates

transcript A-B-D after cocaine.

H3K4me1

H3K4me3

A AB BC D DH3K27Ac

5hmC

Cocaine

AAAA

5hmC

Tet1

AAAA

M a r i n a C o r r a l S p e n c e / N a t u r e P u b l i s h i n g G r o u p


3/3

480 VOLUME 18 | NUMBER 4 | APRIL 2015 NATURE NEUROSCIENCE


neuroscience, computational science and, of

course, psychology (Fig. 1). These fields haveapproached this issue in different ways and

each can inform and motivate future direc-tions in motor control.

In psychology, reward and punishmenthave long been recognized as instrumental for

learning. As early as 1898, Edward Thorndike’slaw of effect stated that if a response leads to a

“satisfying state of affairs” it will be strength-ened and, conversely, if it leads to unpleasant

consequences it will be weakened4. The thesis

that reward is a better motivator than pun-

ishment was also at the core of B.F. Skinner’sprinciple of reinforcement. Operant condi-tioning developed systematic reinforcement

schedules to enhance learning and therebyshape behavior5. However, social psychologists

have also reminded us that human nature isfar more nuanced and more than a collection

of systematically reinforced associations. Manystudies have highlighted the mediating effects

of emotions, such as threat, anxiety, pride orshame, on behavior. Invoking stereotypes, such

as inferior performance of females in math-ematics or athletics, lowers test performance.

The authors build on previous studies that

have shown the crucial importance of rewardon retention2,3, but they now contrast the

effect of reward with that of punishment bydifferentiating their effects on acquisition rate

and retention. This study examined partici-pants moving a cursor to targets displayed on

a screen, steering with their hand movements

hidden from view. To create a learning chal-lenge, the cursor position was rotated by 30

degrees and participants had to practice tosuccessfully reach the target. This exercise is

similar to moving a computer mouse when you

turn it upside down: a challenge that one mas-ters with practice. The specific question of thispaper was how money received for good per-

formance (reward) or lost for bad performance(punishment) would affect the rate of learning

and the retention of the acquired performance.The authors found that punishment acceler-

ated the rate of adaptation, whereas rewardimproved retention of the new mapping.

This study is at the crossroads of at leastthree research disciplines that have exam-

ined the consequences of motivational anderror feedback on motor performance:

the usage of specific splice sites. Specifical ly,alternative splice isoforms upregulated after

cocaine were more likely to be associatedwith increased 5hmC at the corresponding

splice site, whereas 5hmC at splice siteswas more likely to be reduced for isoforms

downregulated after cocaine (Fig. 1).These data provide functional evidence

that 5hmC may regulate splice site usage,

which will be an important area for futureinvestigation.

The last question the authors asked iswhether changes in 5hmC contribute to the

persistent changes in NAc physiology thatare both induced by chronic cocaine and

relevant to addiction. The authors founda significant global correlation between

genes that showed increased 5hmC follow-ing repeated cocaine and those that showed

enhanced steady-state expression 24 h afterwithdrawal from chronic cocaine. The corre-

lation with increased 5hmC was even stron-ger for genes that were induced by a cocaine

challenge. These data therefore indicate

that 5hmC levels not only reflect currenttranscriptional states, but also predict the

potential for genes to turn on in response toa future stimulus. Final ly, the authors dem-

onstrated that, at least for a subset of genes,both mRNA induction and cocaine-induced

changes in 5hmC can persist for at least 1month after the cessation of cocaine expo-

sure. Thus, rather than being just an inter-mediate in the demethylation of DNA, these

data support a model of 5hmC as a meaning-ful epigenetic mark of its own, with potential

functions in the maintenance of transcrip-tional memory.

This work by Feng et al.1 underscores theimportance of epigenetic mechanisms of chro-

matin regulation in the long-lasting changesin neuronal gene expression that are induced

by chronic cocaine. Furthermore, their find-ings demonstrate the power of genome-level

sequencing techniques to open new windows

of understanding into the mechanisms ofneuronal adaptation. The challenge for thefuture will be to distill the detailed chromatin

landscape revealed here into a set of principles

for gene regulation that will better linkmolecular mechanism via cellular function to

the maladaptive circuit changes that underliedrug addiction.

COMPETING FINANCIAL INTERESTS

The author declares no competing financial interests.

1. Feng, J. et al. Nat. Neurosci. 18, 536–544(2015).

2. Lister, R. et al. Science 341, 1237905 (2013).3. Kriaucionis, S. & Heintz, N. Science 324, 929–930

(2009).4. Tahiliani, M. et al. Science 324, 930–935

(2009).5. Koh, K.P. et al. Cell Stem Cell 8, 200–213 (2011).6. Ito, S. et al. Nature 466, 1129–1133 (2010).7. Dawlaty, M.M. et al. Dev. Cell 24, 310–323

(2013).8. Szulwach, K.E. et al. Nat. Neurosci. 14, 1607–1616

(2011).9. Rudenko, A. et al. Neuron 79, 1109–1122

(2013).10. Li, X. et al. Proc. Natl. Acad. Sci. USA 111,

7120–7125 (2014).11. Kaas, G.A. et al. Neuron 79, 1086–1093(2013).

12. Wen, L. et al. Genome Biol. 15, R49 (2014).

Dagmar Sternad is in the Departments of Biology,

Electrical and Computer Engineering, and Physics,

and the Center for the Interdisciplinary Research on

Complex Systems, Northeastern University, Boston,

Massachusetts, USA, and Konrad Paul Körding

is in the Sensory Motor Performance Program,

Rehabilitation Institute of Chicago, Chicago, Illinois,

USA, and the Departments of Physical Medicine

and Rehabilitation, and Physiology, Northwestern

University, Chicago, Illinois, USA.

e-mail:[email protected] [email protected]

Carrot or stick in motor learning

Dagmar Sternad & Konrad Paul Körding

A study shows that reward and punishment have distinct influences on motor adaptation. Punishing mistakes

accelerates adaptation, whereas rewarding good behavior improves retention.

We both love salsa dancing, but learning

salsa is not easy. When one partner missesa step, the other may punish him with a

frown, but when he masters a new move, herpraise rewards him—or does it make him

complacent? Carrot or stick: the mannerby which reward and punishment affects

motor learning is a long-standing questionin education, sports, therapy and beyond.

In this issue of Nature Neuroscience, Galea

et al.1 address this question using a sim-ple reaching task in a perturbed visual

environment.

cocaine and chromatin

Documents