Circuit-dependent striatal PKA and ERK signalingunderlies rapid behavioral shift in mating reaction ofmale miceAkihiro Gotoa,b,c, Ichiro Nakaharaa,d, Takashi Yamaguchia, Yuji Kamiokae, Kenta Sumiyamaf, Michiyuki Matsudab,e,Shigetada Nakanishia,1, and Kazuo Funabikia,d,1

aDepartment of Systems Biology, Osaka Bioscience Institute, Osaka 565-0874, Japan; bLaboratory of Bioimaging and Cell Signaling and dLaboratory ofMolecular Cell Biology and Development, Graduate School of Biostudies, and eDepartment of Pathology and Biology of Diseases, Graduate School of Medicine,Kyoto University, Kyoto 606-8507, Japan; cLaboratory for Memory Mechanisms, Brain Science Institute, RIKEN, Saitama 351-0198, Japan; and fLaboratory forMouse Genetic Engineering, Quantitative Biology Center, Cell Design Research Core, RIKEN, Hyogo 650-0047, Japan

Contributed by Shigetada Nakanishi, April 17, 2015 (sent for review March 18, 2015; reviewed by Kazuto Kobayashi and Masayuki Masu)

The selection of reward-seeking and aversive behaviors is con-trolled by two distinct D1 and D2 receptor-expressing striatal mediumspiny neurons, namely the direct pathway MSNs (dMSNs) and theindirect pathway MSNs (iMSNs), but the dynamic modulation ofsignaling cascades of dMSNs and iMSNs in behaving animals remainslargely elusive. We developed an in vivo methodology to monitorFörster resonance energy transfer (FRET) of the activities of PKA andERK in either dMSNs or iMSNs by microendoscopy in freely movingmice. PKA and ERK were coordinately but oppositely regulated be-tween dMSNs and iMSNs by rewarding cocaine administration andaversive electric shocks. Notably, the activities of PKA and ERK rapidlyshifted when male mice became active or indifferent toward femalemice during mating behavior. Importantly, manipulation of PKA cas-cades by the Designer Receptor recapitulated active and indifferentmating behaviors, indicating a causal linkage of a dynamic activityshift of PKA and ERK between dMSNs and iMSNs in action selection.

in vivo FRET imaging | microendoscope | dorsal striatum | action selection |mating behavior

In changing environments, animals are forced to choose actionsamong several alternatives to survive and to keep offspring for

the next generation. A brain region critical for initiation andselection of actions is the striatum (1–3), and its dysfunctionleads to devastating neurological and psychiatric disorders such asParkinson’s disease and drug addiction (4–7). The striatum receivesconvergent glutamatergic inputs from virtually all cortical areas andthe thalamus, and dopaminergic inputs from the substantia nigrapars compacta (SNc) and the ventral tegmental area (8, 9). Theglutamatergic inputs convey various sensory, motor, and cognitiveinformation and drive striatal medium spiny projection neurons(MSNs) to fire, whereas the dopaminergic inputs strongly influencesynaptic transmission and excitability of MSNs (10, 11). MSNs aredivided into two subpopulations, the direct pathway MSN (dMSN)expressing Gs-coupled dopamine D1 receptors and sending axonsdirectly to the internal segment of the globus pallidus (GPi) and thesubstantia nigra pars reticulata (SNr), and the indirect pathwayMSN (iMSN) expressing Gi-coupled D2 receptors and sendingaxons indirectly to the SNr via the external segment of the globuspallidus (GPe) and subthalamic nucleus (12, 13).Intracellular signaling cascades operating in these two types of

MSNs are important for plastic modification of synaptic trans-mission and excitability (14, 15). Among them, protein kinase A(PKA) and extracellular signal-regulated kinase (ERK) have beenshown to be the key molecules in these signaling cascades (10, 14,16). PKA is positively and negatively regulated by Gs-coupled D1and Gi-coupled D2 receptors, respectively, and contributes to theregulation of a wide range of cellular substrates (10, 11, 16). ERKhas been shown to sense coincidental dopaminergic and gluta-matergic activations to induce protein synthesis and synapticmodification (17).

Although both the ventral and dorsal striatum are essential fornaturally occurring reward-seeking and aversive behaviors, such aseating, mating, and escaping from uncomfortable environments(7, 18, 19), the dorsal striatum plays a key role in efficientlychoosing suitable outcomes of actions (1, 2). Among naturallyoccurring rewarding behaviors, male mating behavior is particularlyinteresting, because it represents innate rapid selection behavior,sometimes actively seeking a female animal but at other times be-coming impassive toward it (20, 21). However, little is known abouthow dMSNs and iMSNs are involved in such action selection. Thepoor understanding of dynamic intracellular signaling mechanismsin action selection behavior is mainly due to the lack of effectivetechniques to monitor time-lapsed changes in such activities inbehaving animals. In this study, we developed an in vivo method-ology in which genetically encoded Förster resonance energytransfer (FRET) biosensors of PKA and ERK (22) were specificallyexpressed in either dMSNs (D1-PKA and D1-ERK) or iMSNs (D2-PKA and D2-ERK), and cell-specific fluorescence changes in theactive and inactive forms of PKA and ERK of the dorsal striatumwere monitored by newly designed microendoscopy in freely mov-ing mice (23, 24). We revealed that the activities of PKA and ERKin dMSNs and iMSNs rapidly shifted when male mice became ac-tive or impassive toward a female mouse during mating behaviorand that the dynamic activity shift of PKA and ERK between

Significance

Selection of actions that allow the seeking of rewards andavoidance of uncomfortable environments is a fundamental an-imal behavior. Here, we report an in vivo method, in which theactivities of PKA and ERK were optically recorded by micro-endoscopy of Förster resonance energy transfer responses ofbiosensors in distinct D1 and D2 dopamine receptor-expressingneurons of the dorsal striatum. The PKA and ERK were co-ordinately but reciprocally regulated not only by rewarding andaversive stimuli but also between the two parallel projectionneurons. Importantly, the cell type-specific regulation of PKA andERK was causally linked to active and indifferent mating re-actions of male mice. The dynamic modulation of PKA and ERK inthe striatum underlies the selection of alternative actions.

Author contributions: A.G., M.M., S.N., and K.F. designed research; A.G. and K.F. performedresearch; I.N., Y.K., K.S., and M.M. contributed new reagents/analytic tools; I.N. developedsoftware to collect and analyze data; T.Y. performed histological analysis; A.G. and T.Y.analyzed data; and A.G., M.M., S.N., and K.F. wrote the paper.

Reviewers: K.K., Fukushima Medical University; and M.M., University of Tsukuba.

The authors declare no conflict of interest.1To whom correspondence may be addressed. Email: funabiki@ent.kuhp.kyoto-u.ac.jp orsnakanis@obi.or.jp.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1507121112/-/DCSupplemental.

6718–6723 | PNAS | May 26, 2015 | vol. 112 | no. 21 www.pnas.org/cgi/doi/10.1073/pnas.1507121112

Dow

nloa

ded

by g

uest

on

June

30,

202

0

dMSNs and iMSNs was causally linked to the selectionof alternative actions.

ResultsDevelopment of FRET Biosensor-Expressing Transgenic Mice and FRETMicroendoscopy. We previously established a method to generatetransgenic mice expressing FRET biosensors, with which the timelapse of activity changes in protein kinases can be imaged (25, 26).To achieve dMSN- and iMSN-specific labeling of the PKA andERK FRET biosensors, we created two flox lines of transgenicmice, floxed-AKAR3EV (PKA) and floxed-EKAREV (ERK)(27), both of which contained a nuclear export signal to localizethese kinases in the cytoplasm (Fig. 1A). Crossing of the two floxlines with D1-Cre and D2-Cre BAC transgenic mice (28) generatedfour species of transgenic mice: D1-PKA mice expressingAKAR3EV in dMSNs, D1-ERK mice expressing EKAREV indMSNs, D2-PKA mice expressing AKAR3EV in iMSNs, and D2-ERK mice expressing EKAREV in iMSNs. We confirmed that theFRET biosensors in D1-PKA and D1-ERK mice were specificallyexpressed in dMSNs whose axons projected to the two main outputnuclei of the basal ganglia: the GPi and the SNr (Fig. S1A). In D2-PKA and D2-ERK mice, the FRET biosensors were selectivelyexpressed in iMSNs whose axons terminated in the GPe (Fig. S1A).FRET imaging of the striatum of freely moving mice was

achieved by developing a fiber bundle-based microendoscope tech-nique and using a confocal laser scanning microscope equippedwith a 445-nm laser for excitation and a pair of band-path filters,483 ± 16 nm for CFP and 542 ± 13 nm for FRET (Fig. 1B, Fig. S2,and Movie S1). On insertion of the endoscopic probe into thebrain, excitation of the FRET biosensors was exclusively detected

in the striatum but not in the cerebral cortex covering the striatum(Fig. S1B). When FRET responses (changes in the FRET/CFPratio) were calculated for the individual cellular structures of the D1-PKA mice by subtracting background levels, the increase in FRETresponses by application of a cAMP analog (8-bromoadenosine-3′,5′-cyclic monophosphorothioate, Sp isomer—Sp-8-Br-cAMPS)through a cannula attached to the side of the endoscope exceeded20% and was comparable to that reported for cultured cells (22, 29)(Fig. 1 C and D and Fig. S3). Furthermore, the FRET responsescalculated for an entire field of view corresponded well to the av-eraged FRET responses of each cellular structure (Fig. 1E). So thissimple and practical approach to measure FRET responses in theentire field of view was adopted for quantification of FRET exci-tation in subsequent analyses.In vivo administration of Sp-8-Br-cAMPS progressively in-

creased FRET responses in both D1-PKA and D2-PKA mice (Fig.1F, Left), but not in the striatum expressing phosphorylationsite-mutated, FRET-negative PKA biosensors (Fig. S3D). Con-versely, the MEK inhibitor 2-(2-chloro-4-iodophenylamino)-N-cyclopropylmethoxy-3, 4-difluorobenzamide (PD184352) graduallydecreased FRET responses in both D1-ERK and D2-ERK mice(Fig. 1F, Right). These results showed that activities of PKA andERK could be monitored by the FRET microendoscopic observa-tion. No alteration in cell number, shape, or striatal organizationwas detected by direct visualization (Movie S1) or histologicalanalysis with DAPI staining and anti-GFP and antityrosinehydroxylase immunostaining of the endoscope-implanted striatum(Fig. S2F).

Sp-8-Br-cAMPS (9)Saline (7)

Sp-8-Br-cAMPS (5)

FRE

T re

spon

se

FRE

T re

spon

seFR

ET

resp

onse

FRE

T re

spon

se

Time (min) Time (min)

Time (min) Time (min)10 20 30 40 50 60

D1-

PKA

D2-

PKA

PD184352 (10)Saline (6)

PD184352 (6)Saline (6)

-2

-1

0

1

0

1

0 10 20 30 40-10

D1-

ERK

D2-

ERK

Saline (8)

Expression of tdKeima

Expression of FRET biosensor

+Stop

Stop

tdKeimaloxP loxPCAG

FRET biosensor

StopFRET biosensor

CFP YFP Sensor Lligand

CFP YFP Sensor Lligand Cre loxP

loxP

CAG

tdKeima

Cre 445 nm laser

2~4 m

f=75 mmKinematicmirror

CFP

FRETNA=0.4

~~

Scanner

Pointed end Flat endA B

F

D E

C

FRE

T re

spon

se

FRE

T re

spon

se

FRE

T re

spon

se

-10

10

30

40

20

0

-10 0 10 20 30 40 50 -10 0 10 20 30 40 50-1

0

0.5

1.0

2.0

1.5

-10

0

10

20AveragedCells 1~10

Cells averagedOverall field

Field without cells

1

2 3

45

6

7

8

910

1

2 3

45

6

7

8

910

0 min 50 minSp-8-Br-cAMPS

D

1-PK

A

FRET

/CFP

1.5

0.120 µm

300 µm 300 µm

(

)

(

)

Fig. 1. Live imaging of PKA and ERK activities in dMSNs and iMSNs of transgenic mice expressing FRET biosensors. (A) Construct for Cre-dependent ex-pression of FRET biosensors. (B) Schema of the microendoscope system. (C) A representative microendoscopic image of FRET responses of 10 cellular structuresof D1-PKA mice. (D) Changes in FRET responses of 10 cellular structures were calculated by subtracting background fluorescence levels and averaged (red).The black bar indicates the local application of Sp-8-Br-cAMPS (2 mM). (E) Changes in FRET responses of an entire field of view (blue) as calculated withoutsubtracting background fluorescence (black) paralleled those of the averaged responses of individual cells (red). (F) FRET responses to local application (black bars)of Sp-8-Br-cAMPS (2 mM) and PD184352 (10 μM) into the dorsal striatum of anesthetized mice. In this and other figures, error bars (vertical lines) show SEM; deepcolors in traces indicate statistically significant changes (P < 0.05; Mann–Whitney U test) from 0 min. The numbers of animals analyzed are given in parentheses.

Goto et al. PNAS | May 26, 2015 | vol. 112 | no. 21 | 6719

NEU

ROSC

IENCE

Dow

nloa

ded

by g

uest

on

June

30,

202

0

PKA and ERK Responses of dMSNs and iMSNs to Cocaine Injection. Toexplore molecular activities underlying rewarding action, wemonitored activities of PKA and ERK in dMSNs and iMSNsupon cocaine injection. This stimulus is known to activate PKAand ERK of dMSNs and to induce hyperlocomotion by blockingdopamine uptake (30). Upon cocaine administration (25 mg/kg,i.p.), the PKA and ERK activities in dMSNs rapidly increasedconcomitantly with the increased locomotion (Fig. 2A). In-terestingly, this increase in PKA and ERK activities continued upto at least 60 min after cocaine administration. Surprisingly, uponcocaine administration, both PKA and ERK activities decreasedin the iMSNs (Fig. 2B). Previous immunohistochemical analysesdid not find such a decrease in the activities of PKA and ERK(30), implying that the FRET biosensors were more sensitive,affording detection of a decrease from the baseline activities. Inthis series of experiments, cocaine was administered by i.p. in-jection. As a control for this cocaine administration, we injectedsaline without cocaine. Both D2-PKA and D2-ERK activitiesshowed a significant increase in response to saline injection. Themesencephalic dopaminergic neurons are known to show transientsuppression of firing in response to the aversive stimuli (2, 31), andthis suppression of dopaminergic inputs supposedly leads to theactivation of PKA and ERK in iMSNs through Gi-coupled D2receptors. We thus hypothesized that the procedure of i.p. in-jection may have served as an aversive stimulus to induce D2-PKAand D2-ERK activation in iMSNs. We tested this hypothesis moredirectly by monitoring striatal PKA and ERK activities in responseto aversive footshock stimuli (Fig. 3).

PKA and ERK Responses of dMSNs and iMSNs Against AversiveFootshock Stimuli. As for aversive reactions to electric foot-shock, we measured the time spent exhibiting a self-grooming be-havior, which is a typical behavioral readout of an aversive reaction(32, 33). We then examined how PKA and ERK responses werecorrelated with the duration of self-grooming after electric foot-shocks. Electric footshocks (2 mA, 40 Hz for 2 s) markedly activated

D2-PKA and D2-ERK and inactivated D1-PKA and D1-ERK (Fig.3A, upper red), together with the concomitant emergence of a self-grooming behavior (Fig. 3A, lower red). Self-grooming was neverobserved in those mice not receiving the electric footshocks. Ingeneral, the duration of PKA and ERK activation and inactivationwas longer than that of the self-grooming behavior.In addition to an acute aversive reaction against a single

electric footshock, we also addressed whether PKA and ERK couldbe involved in the induction of aversive learning by pairing electricshocks with the sound of a bell (Fig. 3A, blue). After habituationwith repeated exposure to a bell sound (2 s in duration), an addi-tional bell sound neither influenced the activities of PKA and ERKin dMSNs and iMSNs nor evoked a self-grooming behavior, in-dicating that a bell sound per se served as neutral information (Fig.3A, gray). Notably, when the bell sound was conditioned withelectric footshocks, this conditioned stimulus induced both PKAand ERK activation and inactivation in iMSNs and dMSNs,respectively, together with the concomitant expression of self-grooming behavior (Fig. 3A, blue). Importantly, the sum of durationexhibiting self-grooming behavior was correlated positively with theD2-PKA and D2-ERK activities and negatively with the D1-PKAand D1-ERK activities, respectively (Fig. 3B). Saline injection,

FRE

T re

spon

se

FRE

T re

spon

seFR

ET

resp

onse

FRE

T re

spon

seD

2-P

KA

D2-

ER

K

D1-

PK

A

D1-

ER

K

A

B

Fig. 2. Reciprocal regulation of PKA and ERK activities between dMSNs andiMSNs in response to cocaine or saline injection. PKA and ERK activities ofdMSNs (A) and iMSNs (B) (upper traces) and locomotor activities (lower traces)in response to cocaine (red) or saline (gray) injection. There was no significantdifference in locomotor activities in response to cocaine or saline injectionamong four lines of transgenic mice (P > 0.05; Mann–Whitney U test).

D1-

PK

AD

2-P

KA

D2-

ER

KD

1-E

RK

FRE

T re

spon

se (%

)

**r=-0.682

*r= 0.636 **r= 0.779

Self-grooming (sec) Self-grooming (sec) Self-grooming (sec) Self-grooming (sec)

D1-PKA D2-PKA D2-ERK

-1.6

-1.2

-0.8

-0.4

0

0 200 4000.2

0.6

1.0

1.4

0 200 4000.2

0.6

1.0

1.4

0 200 400

A

B

400

**r=-0.662

D1-ERK

-1.6

-1.2

-0.8

-0.4

0

0 200

Fig. 3. Reciprocal regulation of PKA and ERK activities between dMSNs andiMSNs in response to electric footshocks. (A) PKA and ERK activities (uppertraces) and aversive self-grooming behavior (lower traces) in response toelectric footshocks (Shock), bell sound without unconditioned stimulus(nonconditioned stimulus, NCS), and bell sound after conditioning withelectric shocks (conditioned stimulus, CS). The time spent exhibiting self-grooming was measured every 3 min. The vertical dashed lines indicate astarting time point of these stimuli. (B) Average changes in FRET responses ofPKA and ERK of dMSNs and iMSNs were plotted against the time spentexhibiting self-grooming behavior during the 30-min period. Black and graycircles represent electric shock and conditioned bell sound data, respectively,taken from A. The r value was calculated by use of Pearson correlation analysis;*P < 0.05, **P < 0.01.

6720 | www.pnas.org/cgi/doi/10.1073/pnas.1507121112 Goto et al.

Dow

nloa

ded

by g

uest

on

June

30,

202

0

which mildly activated D2-PKA and D2-ERK (Fig. 2B), also in-duced less but significant levels of self-grooming behavior (Figs. S4and S5). However, there was no difference in extents of self-grooming behavior among the four lines of transgenic mice (Fig.S5). These results thus indicated a tight linkage between the D2-PKA and D2-ERK activation and the expression of aversive be-haviors. Noteworthy also is that transient self-grooming wasevoked immediately after cocaine injection but disappearedthereafter (Fig. S4). This finding strongly suggested that theopposing effects of cocaine administration on the PKA andERK signaling cascades counteracted the aversive reaction in-duced by i.p. injection. Collectively, the FRET microendoscopyanalysis thus allowed us to monitor dynamic and cell type-specific changes in the activities of PKA and ERK and demon-strated that the activities of PKA and ERK were reciprocallycontrolled not only by rewarding and aversive reactions but alsobetween dMSNs and iMSNs.

PKA and ERK Responses in Male Mating Behavior. Mating reactionsrepresenting a naturally occurring reward-seeking behavior (18,34) have been shown to induce an increase in dopamine con-centrations in both ventral and dorsal striatum (35). Themating reactions were quantified by determining the percentagesof time in which the male mouse exhibited characteristic matingbehaviors (sniffing, grooming, and mounting) toward a femalemouse (Movies S2–S9) (34) every 1 min during a 30-min period

(% positive mating reaction, % PMR). A significant positive corre-lation was noted between the extent of D1-PKA and D1-ERK ac-tivation and % PMR (Fig. 4A), whereas a negative correlation wasfound in the case of D2-PKA and D2-ERK (Fig. 4A). When matingreactions were arbitrarily divided into frequent and infrequent in-teractions by a value of 20% PMR (Fig. 4B), the D1-PKA and D1-ERK activities significantly increased in the frequently interactingmice but not in the infrequently interacting ones. In clear contrast,the D2-PKA and D2-ERK activities increased in the infrequentlyinteracting ones (Fig. 4B). Notably, there was no significant differ-ence in locomotive distance between frequent and infrequent in-teraction groups (Fig. 4B, Bottom), suggesting that the mice in theinfrequent interaction group avoided or escaped from the femalewith an extent of locomotion comparable to that of the male mice inthe frequent interaction group that chased the female. These resultsindicated that the activation of PKA and ERK of iMSNs was alsocritically involved in impassive and avoiding behavior toward afemale mouse.We also noticed that the successive active and indifferent/

avoiding phases of a male mouse in mating behavior resulted in thefrequent and infrequent mating reactions. We thus more carefullypursued temporal changes in the PKA and ERK activities duringthe mating reaction of individual mice. The rapid changes in activeor indifferent phases of mating reactions were defined as more thana 50% PMR increase or decrease within 1–4 min, respectively (Fig.5 A and B and Fig. S6). Fig. 5A illustrated four representative miceindicating close association of PKA activity with mating behavior.D1-PKA mouse 1 started sniffing immediately after female entryand showed a rapid concomitant increase in PKA activity. D1-PKAmouse 2 did not interact with the female mouse and showed noobvious increase in PKA activity. D2-PKA mouse 1 showed astrong interest toward the female mouse during the first 4 min butstopped PMR thereafter (Movie S8). The D2-PKA activity of thismouse concomitantly increased and remained high during the in-different phase of mating reaction. D2-PKA mouse 2 showed nointerest for the first 17 min and then started sniffing and groomingthereafter (Movie S9). Concomitantly, the D2-PKA activity shiftedfrom activation to inactivation. Statistical analysis of the rapidchange of PKA and ERK activities during mating behaviors ex-plicitly indicated that the activities of D1-PKA and D1-ERK weremarkedly elevated during the active phase, whereas those of D2-PKA and D2-ERK were increased when a male mouse becameindifferent to or escaped from a female mouse (Fig. 5B and Fig.S6). Thus, rapid change in the activities of both PKA and ERKreflected a shift between active and indifferent phases of themating reaction. To substantiate this observation, we paired amale mouse with a hormonally primed female mouse to facilitateejaculation by the male mouse (Fig. 5 C and D and Movie S10).The D1-PKA and D1-ERK activities were elevated during themating reaction, but the D1-PKA activity was rapidly and signifi-cantly reduced to lower levels within 4 min after ejaculation,suggesting that PKA in the dMSNs was closely associated with therapid shift in the mating reaction.

Pathway-Specific PKA Modification by DREADD and Its Effect onMating Behavior. Finally, we examined the causal linkage of thePKA signaling cascade in induction and suppression of matingreactions by activation of Gs and Gi specific for either dMSNs oriMSNs by using the Designer Receptors Exclusively Activated by aDesigner Drug (DREADD) method (36). We used two lines of floxadeno-associated viruses (AAVs) in which Gs-activating rM3Ds-mCherry and Gi-activating hM4Di-mCherry were bounded by loxPsequences (Fig. S7A). The transfection of D1- or D2-PKA micewith the AAV–rM3Ds-mCherry or the AAV–hM4Di-mCherryresulted in specific and exclusive expression of the respective de-signer receptors in either dMSNs or iMSNs of the transgenic mice(Fig. S7B). In the experiments using Gi-activated D1-PKA andGs-activated D2-PKA male mice, we paired a male mouse with a

Female entry

Removal

0

50

100

050

100

0.0

0.5

0.0

0.5

-0.5

1.0

1.5

0.0

0.5

1.0

0.0

0.5

-0.5

1.0

1.5

0

50

100

FRE

T re

spon

se

Time (min) Time (min)

D2-

PKA

D2-

ERK

D1-

ERK

D1-

PKA

D1-PKA

FRE

T re

spon

se

B

A D2-PKAD1-ERK D2-ERK

FRE

T re

spon

seFR

ET

resp

onse

FRE

T re

spon

se

Fig. 4. PKA and ERK activities of dMSNs and iMSNs during sexual in-teraction. (A) A male mouse was exposed to an unfamiliar female mouse for30 min. Average changes in FRET responses of PKA and ERK during the30-min period were plotted against % PMR. The r value was calculated byuse of Pearson correlation analysis; *P < 0.05. (B) Changes in activities of PKAand ERK (upper traces) and locomotor activities (lower traces) of frequentlyor infrequently interacting male mice.

Goto et al. PNAS | May 26, 2015 | vol. 112 | no. 21 | 6721

NEU

ROSC

IENCE

Dow

nloa

ded

by g

uest

on

June

30,

202

0

hormonally primed BALB/c female mouse to increase the basallevel of mating reactions of the male mouse (21). In the experimentsusing Gs-activated D1-PKA and Gi-activated D2-PKA male mice,we used a large-sized ICR female mouse to decrease the basal levelof the mating reaction of the male mouse. This procedure helped inclarifying the effect of the cell-specific activation of Gs and Gi in themale mating reaction. When the designer receptor agonist cloza-pine-N-oxide (CNO) was injected 10–20 min after the female entry,this treatment considerably decreased FRET responses of PKA inthe hM4Di-expressing MSNs of both D1-PKA and D2-PKA mice(Fig. 6, Left). In accordance with this decrease in PKA activity,these mice exhibited significant suppression and enhancement,respectively, of % PMR 10–20 min after CNO treatment (Fig. 6,Left). The effect of CNO treatment was reversed in the rM3Ds-expressing D1-PKA and D2-PKA mice (Fig. 6, Right). The in-crease in PKA activity in the D1-PKA and the D2-PKA micesignificantly enhanced and suppressed, respectively, the % PMRof the corresponding PKA mice, although the enhancement of %PMR of the Gs-activated D1-PKA mice was transient after CNOtreatment. Importantly, the CNO effect was specific for the matingreaction, as no difference in locomotor activity was detected

following CNO treatment in the four lines of transgenic mice.These results indicated a causal linkage between the dorsal striatalPKA activity of dMSNs and iMSNs and the mating reaction ofmale mice.

DiscussionThis study demonstrated that PKA and ERK were coordinatelybut oppositely regulated between dMSNs and iMSNs in thedorsal striatum under different behavioral conditions. A sensorystimulus with bell sound had no effect on the PKA and ERKactivities of either dMSNs or iMSNs of mice after habituationwith a repeated bell sound, but rewarding and aversive stimulimarkedly and reciprocally modulated the activities of both ki-nases in dMSNs and iMSNs. Animal behaviors can thus be or-derly aligned from rewarding to aversive/impassive behaviors bythe extent of activation and inactivation of PKA and ERK (Fig.S8). Remarkably, the rapid shift in the PKA and ERK activitiesof both dMSNs and iMSNs was tightly associated with the activeand indifferent phases of naturally occurring mating reactions,indicating that reciprocal but highly synergistic activation andinactivation of the PKA and ERK cascades in dMSNs andiMSNs play a key role in selection of actions in the male matingbehavior. This conclusion was verified by use of the DREADD

Indifferent phaseActive phase

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

Female entry

D1-PKAD1-ERK

D2-PKAD2-ERK

D2-PKAFR

ET

resp

onse

0.0

0.5

-10 0 200

50

100

D1-PKAmouse 1 mouse 2 mouse 1 mouse 2

D1-PKA

Time (min)Time (min) Time (min) Time (min)-10 0 10-10 0 10

0.0

0.5

1.0

0

50

100

0

50

100

FRE

T re

spon

se

D2-PKA

0 20-10

-0.5

0.0

0.5

0

50

100

A

B

Time (min)

C

D

D1-

PKA

D1-

ERK

FRE

T re

spon

se

FRE

T re

spon

seFR

ET

resp

onse

Female entry

EjaculationD1-ERK

D1-PKA

Fig. 5. PKA and ERK responses of dMSNs and iMSNs upon a rapid shift inmating behavior. (A) Four representative mice indicating close association ofPKA activity (upper traces) with mating behavior (lower traces). The red andblue lines in the lower traces show a rapid shift between the active andindifferent phases of mating reaction, which were defined as more than a50% increase and decrease, respectively, in % PMR within 1–4 min. (B) Up-and down-regulation of PKA and ERK activities at the active and indifferentphases of the mating reaction is summarized from the data of Fig. S6 (*P <0.05; Mann–Whitney U test). The numbers in parentheses indicate thesample numbers from six to nine animals. (C) Examples of changes in thePKA and ERK activities in dMSNs after ejaculation. Ejaculation is marked withthe arrow in the D1-PKA (red) and D1-ERK (black) mouse. (D) Rapid in-activation of PKA in the D1-PKA mice after ejaculation (*P < 0.05; n = 5;Wilcoxon signed-rank test).

FRE

T re

spon

seFR

ET

resp

onse

D1-

PKA

FRE

T re

spon

seD

1-PK

A

Gi (AAV-hM4Di-mCherry) Gs (AAV-rM3Ds-mCherry)

Gi (AAV-hM4Di-mCherry) Gs (AAV-rM3Ds-mCherry)

Female entry

D2-

PKA

FRE

T re

spon

seD

2-PK

A

5025

75

0

5025

75

0

5025

75

0

5025

75

0

Fig. 6. PKA response and change in mating reactions induced by Gi or Gsactivation by DREADD. Gi (Left) and Gs (Right) of the D1-PKA and D2-PKAmice were activated by CNO injection 10–20 min after female entry. Upper,middle, and lower traces show FRET response, % PMR, and locomotive dis-tance, respectively. The numbers in parentheses indicate the sample num-bers in which FRET responses were measured by alternately treatingGi-activated D1-PKA (n = 3) and other transgenic mice (n = 4 each) witheither saline or CNO 1 d after each treatment and averaged. Deep red color intraces with CNO injection indicates a statistically significant change (P < 0.05;Mann–Whitney U test) from those with saline injection.

6722 | www.pnas.org/cgi/doi/10.1073/pnas.1507121112 Goto et al.

Dow

nloa

ded

by g

uest

on

June

30,

202

0

technique, in which the activation of Gs and Gi induced activeand impassive mating reactions in an MSN cell type-dependentmanner. Thus, the modulation of the PKA cascade in the dorsalstriatal MSNs is directly linked to the action selection in malemating behaviors.The D2-PKA and D2-ERK in iMSNs were activated by aver-

sive stimuli and in the impassive phase of male mating reaction.This finding is consistent with the predominant expression of high-affinity (nM order) D2 receptors in iMSNs, which are capable ofsensing basal changes in synaptic dopamine concentrations in thestriatum (10, 35, 37). The iMSN transmission would thus greatlycontribute to the innate tendency of animals to be more con-cerned about avoiding uncomfortable and dangerous environ-ments and escaping from a predator. Noteworthy also is that theD2 receptor antagonists are widely used as an effective thera-peutic drug for some psychiatric disorders such as schizophrenia(38). The modulatory disturbance in the D2 receptor signaling iniMSNs may cause an imbalance of mental activity and could resultin possible emergence of some forms of psychiatric disorders.Conversely, when animals encounter rewarding stimuli such assexual interaction, an increase in dopamine levels in the striatumwould reach the threshold for low-affinity (μM order) D1 receptoractivation in dMSNs (10, 35, 37). Although other neurotransmit-ters and striatal interneurons may also be involved in reward-seeking and aversive behaviors (10, 39), the D1 and D2 receptorscould play a key role in distinctly controlling striatum-mediatedaction selection via common but pathway-dependent PKA andERK signaling cascades (37).

The use of deep brain-inserted fiber optics combined withmeasurement of FRET responses of signaling cascades enabledus to explore the regulation of signaling mechanisms in specificsubgroups of MSNs that control action selection. Althoughdetailed analysis of FRET responses of individual cells in be-having animals was not done in this study, multiple optic fibersof the microendoscope could allow investigation of spatio-temporal interactions in the cell assembly of a local circuit.Thus, this methodology could also be used to study regulatorymechanisms of neural circuits involved in a wide range ofanimal behaviors.

Materials and MethodsEssential methodological information needed for a basic understandingof the text has been described in the main text at the appropriate places.SI Materials and Methods contains detailed information on generationof transgenic mice expressing cell type-specific PKA and ERK biosen-sors, fabrication of microendoscope, imaging measurements and anal-ysis by the microendoscope system, and animal behavioral analysesincluding locomotion, cocaine administration, electric footshock, andmating behavior. All animal experiments were approved by the animalresearch committee of Osaka Bioscience Institute.

ACKNOWLEDGMENTS. We thank H. Hioki for the D1-Cre mouse andN. Yamatani for technical assistance. This research was supported by ResearchGrants-in-Aid 22220005 (to S.N.) and 24111552, 22300136, and 26560470 (toK.F.); by Grant-in-Aid for Scientific Research on the Innovative Area “Fluores-cence Live imaging” 22113002 (to M.M.); and by a Grant-in-Aid for JapanSociety for the Promotion of Science Fellows (to A.G.), from the Ministry ofEducation, Culture, Sports, Science, and Technology of Japan.

1. Balleine BW, Delgado MR, Hikosaka O (2007) The role of the dorsal striatum in rewardand decision-making. J Neurosci 27(31):8161–8165.

2. Tai LH, Lee AM, Benavidez N, Bonci A, Wilbrecht L (2012) Transient stimulation ofdistinct subpopulations of striatal neurons mimics changes in action value. Nat Neu-rosci 15(9):1281–1289.

3. Hikida T, Kimura K, Wada N, Funabiki K, Nakanishi S (2010) Distinct roles of synaptictransmission in direct and indirect striatal pathways to reward and aversive behavior.Neuron 66(6):896–907.

4. Albin RL, Young AB, Penney JB (1989) The functional anatomy of basal ganglia dis-orders. Trends Neurosci 12(10):366–375.

5. Chesselet MF, Delfs JM (1996) Basal ganglia and movement disorders: An update.Trends Neurosci 19(10):417–422.

6. Hyman SE, Malenka RC, Nestler EJ (2006) Neural mechanisms of addiction: The role ofreward-related learning and memory. Annu Rev Neurosci 29:565–598.

7. Graybiel AM (2008) Habits, rituals, and the evaluative brain. Annu Rev Neurosci 31:359–387.

8. Grace AA, Floresco SB, Goto Y, Lodge DJ (2007) Regulation of firing of dopaminergicneurons and control of goal-directed behaviors. Trends Neurosci 30(5):220–227.

9. Wilson CJ (2004) The synaptic organization of the brain. Basal Ganglia, ed Shepherd GM(Oxford Univ Press, Oxford), 5th Ed, pp 361–414.

10. Gerfen CR, Surmeier DJ (2011) Modulation of striatal projection systems by dopamine.Annu Rev Neurosci 34:441–466.

11. Tritsch NX, Sabatini BL (2012) Dopaminergic modulation of synaptic transmission incortex and striatum. Neuron 76(1):33–50.

12. Gerfen CR, et al. (1990) D1 and D2 dopamine receptor-regulated gene expression ofstriatonigral and striatopallidal neurons. Science 250(4986):1429–1432.

13. Smith Y, Bevan MD, Shink E, Bolam JP (1998) Microcircuitry of the direct and indirectpathways of the basal ganglia. Neuroscience 86(2):353–387.

14. Girault JA (2012) Signaling in striatal neurons: The phosphoproteins of reward, ad-diction, and dyskinesia. Prog Mol Biol Transl Sci 106:33–62.

15. Shiflett MW, Balleine BW (2011) Molecular substrates of action control in cortico-striatal circuits. Prog Neurobiol 95(1):1–13.

16. Surmeier DJ, Ding J, Day M, Wang Z, Shen W (2007) D1 and D2 dopamine-receptormodulation of striatal glutamatergic signaling in striatal medium spiny neurons.Trends Neurosci 30(5):228–235.

17. Girault JA, Valjent E, Caboche J, Hervé D (2007) ERK2: A logical AND gate critical fordrug-induced plasticity? Curr Opin Pharmacol 7(1):77–85.

18. Salamone JD, Correa M (2012) The mysterious motivational functions of mesolimbicdopamine. Neuron 76(3):470–485.

19. O’Doherty J, et al. (2004) Dissociable roles of ventral and dorsal striatum in in-strumental conditioning. Science 304(5669):452–454.

20. Dickson BJ (2008) Wired for sex: The neurobiology of Drosophila mating decisions.Science 322(5903):904–909.

21. Park JH (2011) Assessment of male sexual behavior in mice. Mood and AnxietyRelated Phenotypes in Mice, ed Gould T (Springer Science, New York), Vol 63, pp357–373.

22. Komatsu N, et al. (2011) Development of an optimized backbone of FRET biosensorsfor kinases and GTPases. Mol Biol Cell 22(23):4647–4656.

23. Vincent P, et al. (2006) Live imaging of neural structure and function by fibredfluorescence microscopy. EMBO Rep 7(11):1154–1161.

24. Flusberg BA, et al. (2005) Fiber-optic fluorescence imaging. Nat Methods 2(12):941–950.

25. Goto A, et al. (2013) GDNF and endothelin 3 regulate migration of enteric neuralcrest-derived cells via protein kinase A and Rac1. J Neurosci 33(11):4901–4912.

26. Kamioka Y, et al. (2012) Live imaging of protein kinase activities in transgenic miceexpressing FRET biosensors. Cell Struct Funct 37(1):65–73.

27. Sumiyama K, Kawakami K, Yagita K (2010) A simple and highly efficient transgenesismethod in mice with the Tol2 transposon system and cytoplasmic microinjection.Genomics 95(5):306–311.

28. Valjent E, Bertran-Gonzalez J, Hervé D, Fisone G, Girault JA (2009) Looking BAC atstriatal signaling: Cell-specific analysis in new transgenic mice. Trends Neurosci 32(10):538–547.

29. Allen MD, Zhang J (2006) Subcellular dynamics of protein kinase A activity visualizedby FRET-based reporters. Biochem Biophys Res Commun 348(2):716–721.

30. Bertran-Gonzalez J, et al. (2008) Opposing patterns of signaling activation in dopa-mine D1 and D2 receptor-expressing striatal neurons in response to cocaine andhaloperidol. J Neurosci 28(22):5671–5685.

31. Ilango A, Shumake J, Wetzel W, Scheich H, Ohl FW (2012) The role of dopamine in thecontext of aversive stimuli with particular reference to acoustically signaled avoid-ance learning. Front Neurosci 6:132.

32. D’Angio M, Serrano A, Driscoll P, Scatton B (1988) Stressful environmental stimuliincrease extracellular DOPAC levels in the prefrontal cortex of hypoemotional (Ro-man high-avoidance) but not hyperemotional (Roman low-avoidance) rats. An in vivovoltammetric study. Brain Res 451(1-2):237–247.

33. Giorgi O, Lecca D, Piras G, Driscoll P, Corda MG (2003) Dissociation between meso-cortical dopamine release and fear-related behaviours in two psychogenetically se-lected lines of rats that differ in coping strategies to aversive conditions. Eur JNeurosci 17(12):2716–2726.

34. Schultz W (2007) Behavioral dopamine signals. Trends Neurosci 30(5):203–210.35. Damsma G, Pfaus JG, Wenkstern D, Phillips AG, Fibiger HC (1992) Sexual behavior

increases dopamine transmission in the nucleus accumbens and striatum of male rats:Comparison with novelty and locomotion. Behav Neurosci 106(1):181–191.

36. Rogan SC, Roth BL (2011) Remote control of neuronal signaling. Pharmacol Rev 63(2):291–315.

37. Nakanishi S, Hikida T, Yawata S (2014) Distinct dopaminergic control of the direct andindirect pathways in reward-based and avoidance learning behaviors. Neuroscience282C:49–59.

38. Seeman P (2010) Dopamine D2 receptors as treatment targets in schizophrenia. ClinSchizophr Relat Psychoses 4(1):56–73.

39. Kreitzer AC, Malenka RC (2008) Striatal plasticity and basal ganglia circuit function.Neuron 60(4):543–554.

Goto et al. PNAS | May 26, 2015 | vol. 112 | no. 21 | 6723

NEU

ROSC

IENCE

Dow

nloa

ded

by g

uest

on

June

30,

202

0