Identification of Optogenetically Activated Striatal Medium Spiny Neurons by Npas4 Expression

Optogenetics is a powerful neuromodulatory tool with many unique advantages to explore functions of neuronal circuits in physiology and diseases. Yet, interpretation of cellular and behavioral responses following in vivo optogenetic manipulation of brain activities in experimental animals often necessitates identification of photoactivated neurons with high spatial resolution. Although tracing expression of immediate early genes (IEGs) provides a convenient approach, neuronal activation is not always followed by specific induction of widely used neuronal activity markers like c-fos, Egr1 and Arc. In this study we performed unilateral optogenetic stimulation of the striatum in freely moving transgenic mice that expressed a channelrhodopsin-2 (ChR2) variant ChR2(C128S) in striatal medium spiny neurons (MSNs). We found that in vivo blue light stimulation significantly altered electrophysiological activity of striatal neurons and animal behaviors. To identify photoactivated neurons we then analyzed IEG expression patterns using in situ hybridization. Upon light illumination an induction of c-fos was not apparent whereas another neuronal IEG Npas4 was robustly induced in MSNs ipsilaterally. Our results demonstrate that tracing Npas4 mRNA expression following in vivo optogenetic modulation can be an effective tool for reliable and sensitive identification of activated MSNs in the mouse striatum.


Introduction
Optogenetic technology is rapidly fostering the study of brain functions through delineation of complex neuronal networks [1]. Incorporation of microbial light-activated regulators of transmembrane conductance in experimental animals allows manipulation of electrical activity of the target tissue with the millisecond temporal precision even in freely moving animals. When functional light-activated channel proteins are expressed in neurons both gain-of-function and loss-of-function studies are possible to modulate animal behaviors and to infer causal relationship between the illuminated neurons and the resultant cellular or behavioral changes in physiological or pathophysiological conditions. For instance, Brown et al. (2010) found that direct optical activation of dopamine (DA) neurons of the mouse ventral tegmental area (VTA) was sufficient to drive redistribution of AMPA receptors [2]. Using an optogenetic approach Deisseroth and colleagues (2009) demonstrated that selective high-frequency stimulation of afferent axons projecting to the subthalamic nucleus robustly ameliorated the disease symptoms in a rodent model of Parkinson's disease (PD) [3].
One of the big issues during in vivo optogenetic manipulation is that the illuminated cells are inevitably heterogeneous in terms of intensity of the incident light [1] and it is difficult to identify photoactivated cells precisely. This limitation calls for additional measures for sensitive detection of photoactivated cells with high spatial resolution. One convenient approach is the visualization of activated neurons by tracing induction of immediate early genes (IEGs) such as c-fos, Egr1 and Arc [4][5][6][7]. Nevertheless, previous studies indicated that c-fos, the most widely used IEG, is not a universal marker for neuronal activation and IEGs may be differentially induced depending on the neuronal population and/ or the stimulus. For instance, Isogai et al. (2011) found that Egr1, but not c-fos, was induced robustly in the mouse vomeronasal organ following sensory stimulation [8]. Though several types of drugs of abuse can induce c-fos in the striatum [9,10], an atypical antipsychotic drug clozapine was found to induce Egr1 but not c-fos mRNAs in the rat striatum [11]. Using in vivo light stimulation followed by in situ hybridization of activity markers here we show that the neuronal IEG Npas4 can identify photoactivation of striatal medium spiny neurons (MSNs) more reliably compared to other commonly used IEGs like c-fos, Egr1 and Arc.

In vivo optical stimulation of striatal MSNs
To drive cell type-specific expression of a highly light-gated channelrhodopsin-2 (ChR2) variant ChR2(C128S) [12] we took advantage of the tetracycline-controlled transcriptional activator (tTA) system. Previously it was demonstrated that tTA is stably expressed in almost all striatal MSNs of PDE10A2-tTA mice [13]. Recently the tetO-ChR2(C128S)-EYFP BAC transgenic mouse line has been established in which the tTA-dependent promoter (tetO) drives the expression of ChR2(C128S) [14]. The above two lines were crossed to achieve striatal MSN-specific strong expression of ChR2(C128S). In the compound heterozygous mice strong fluorescent signals spanned the whole projection area of striatal MSNs including the caudate putamen (CPu), the fibers of striatal MSNs and their projection areas including the external segment of the globus pallidus (GP), the entopeduncular nucleus (EP) and the substantia nigra pars reticulata (SNR) ( Figure 1A). We recorded extracellular neuronal activity of striatal neurons before and after illuminating with a single 100-ms pulse of blue laser (473 nm) in awake mice ( Figure 1B, C). The laser illumination evoked prolonged excitation which lasted after termination of the photostimulation ( Figure 1C). In our experiment, we used only blue light to modulate neuronal activity. Therefore, the excitation after a single pulse of photostimulation lasted more than 1 min (data not shown).
To perform in vivo optogenetic manipulation in freely moving animals we implanted optical fibers in the mouse brain targeting the striatum (Figure 2A). After recovery from implantation procedure mice were kept in their home cages where they did not show any altered behaviors. When awake mice in home cages were given blue light (a single 500-ms pulse) to illuminate the striatum in one hemisphere, animals displayed complex behaviors, including ipsilateral rotations (Sano and Tanaka, manuscript in preparation). These results indicated that light illumination reliably activated the striatal MSNs and the optogenetic modulation of neuronal excitability was sufficient to trigger recognizable behavioral responses in freely moving mice.

Tracing photoactivated neurons
We looked for a selective increase in expression of IEGs in the ipsilateral striatum after 10 minutes of ChR2(C128S)-mediated unilateral photoactivation in the PDE10A2-tTA mice (Figure 2A). Any apparent induction of c-fos, the most widely used activity marker, was not observed in the striatum in our experimental conditions ( Figure 2B1, B1', C). Therefore, it was necessary to check whether the striatum in the ChR2(C128S)-expressing mice can respond normally to known inducers of IEGs. We found that acute administration of methamphetamine (2 mg/kg, intraperitoneal injection, single dose), which is known to induce IEG expression in the striatum [15][16][17][18], significantly upregulated c-fos expression in the striatum of the ChR2(C128S)-expressing transgenic mice and the expression pattern was apparently similar to that of wild type mice ( Figure S1). Additionally, our preliminary observation did not indicate any noticeable behavioral difference between the wild type and the transgenic mice without or with methamphetamine treatment (data not shown). Together, it seemed that the striatal functions were intact in the ChR2(C128S)-expressing PDE10A2-tTA line.
Subsequently, we expanded the screening using a number of IEGs which were found to be induced by brain activities in previous studies. We found strong signals of Arc mRNAs in the illuminated striatum ( Figure 2B3') in addition to very dense signals in other brain regions including the cerebral cortex. Nonetheless, a considerable level of Arc mRNA expression was observed also in the contralateral side ( Figure 2B3, C). Similarly, Junb expression was only slightly higher in the ipsilateral striatum compared to the contralateral striatum (Data not shown). We found that Egr1 expression was not suitable for tracing photoactivation of striatal MSNs since dense mRNA signals appeared in both the control and the illuminated hemispheres ( Figure 2B4, B4', C). In addition, any selective induction was absent for other IEGs such as Fosb, Egr3 and Jun (Data not shown). In stark contrast, another neuronal IEG, the basic helix-loop-helix (bHLH)-PAS transcription factor Npas4, was robustly induced specifically in the ipsilateral striatum following photoactivation of striatal MSNs ( Figure 2B2', C). Signals for Npas4 mRNAs were almost absent in the contralateral sham-treated striatum ( Figure 2B2, C) and in the striatum of the mice in which optical fibers were implanted in both hemispheres but no laser illumination was done ( Figure S2).
We also checked IEG induction with an extended time window, 60 minutes after illumination. In contrast to 10 minutes postillumination delay, there was almost no ISH signal of Npas4 mRNA in the striatum after 60 minutes of photoactivation ( Figure 3). These data were consistent with previous studies in respect to the quick and transient induction of Npas4 [19,20].
In some cases we observed induction of both Npas4 (see later) and c-fos (data not shown) in the ipsilateral cortex after unilateral optical illumination of the striatum. Such cortical IEG induction was absent in the double-instrumented sham-treated mice ( Figure  S2B1-C2). It is possible that the cortical IEG induction we found was secondary effects of striatal stimulation. Previous studies also reported significant induction of c-fos expression in the cortex after striatal stimulation and disinhibition of the thalamocortical pathways has been suggested as a possible mechanism [21][22][23][24]. Although Arc was induced by photostimulation, the expression level was relatively high in the contralateral striatum. Any induction of Egr1 was not apparent after illumination. Data represent mean 6 SEM. ** Difference between groups was highly significant (p#0.01), * Difference between groups was significant (p#0.05). Scale bar: (B) 200 mm. doi:10.1371/journal.pone.0052783.g002 When propagating through a diffuse scattering media like brain tissue the incident light is attenuated resulting from a number of phenomena such as scattering, absorption and conical spreading [25]. Nonetheless, if the target volume is small (,1 mm 3 ) light penetration is not a limiting factor and the entire target could be recruited [3,25]. On the other hand, the mouse striatum is a relatively large structure comprising approximately 5-6% of the brain volume [26] and it extends approximately 4 mm and 3 mm along the rostrocaudal and dorsoventral axes, respectively [27]. Presumably, in vivo optogenetic modulation in mouse striatum will recruit a significant volume of tissue proximal to the tip of the optic fiber and may leave the distant regions mostly unaffected. In our experimental conditions light stimulation was applied to the dorsal striatum ( Figure 4F, G) which has been implicated in motor control [28,29]. Figure 4 shows expression of Npas4 as observed in coronal sections of the brain hemisphere ipsilateral to optical illumination. Npas4 induction took place along the entire mediolateral extent of the striatum near the tip of the optic fiber ( Figure 4F). At the rostral striatum strong induction of Npas4 appeared mostly in the dorsomedial part ( Figure 4B-D) and expression level was low along the ventral striatum ( Figure 4A, B). It is noteworthy that placement of the optical fiber (both location and orientation/angle of the tip) could influence the number of labeled neurons since the neurons which are distant from the probe may not be activated because of attenuation and conical spreading of the incident light [25]. Nevertheless, we observed reliable photostimulation in our experimental set up and it appeared that optical illumination recruited a substantial volume of striatum along the rostrocaudal, mediolateral and dorsoventral axes as revealed by the selective induction of Npas4 mRNA expression (Figure 4, Figure 5A1-A4).
Major components of the basal ganglia neuronal network are MSNs, the most abundant cell type in the striatum, and there are several types of interneurons such as cholinergic interneurons and GABAergic interneurons [30][31][32]. We performed double fluorescent ISH to know which cell types in the striatum expressed Npas4 after optical stimulation. In the illuminated striatum almost all Npas4-positive cells were double-positive for Npas4 and ChR2(C128S)-EYFP ( Figure 5A1-B3). Cholinergic interneurons, also known as tonically active neurons (TANs), exhibit spontaneous firing and are synaptically connected with MSNs [33,34]. We used choline acetyltransferase (Chat) as a marker for cholinergic interneurons. It seemed that in cholinergic interneurons induction of Npas4 was minimal, if any, since none of the Chat positive cells (56 cells examined using confocal microscopy) was co-labeled with Npas4 in the striatum ( Figure 5C1-C3). Seemingly, Npas4 induction was also absent in parvalbumin (Pvalb, also known as PV)-positive GABAergic interneurons as we did not observe any Pvalb and Npas4 double-positive cells (14 cells examined) in the illuminated striatum ( Figure 5D1-D3). Therefore, these data indicated that Npas4 mRNA expression was induced by photostimulation mostly in the striatal MSNs which expressed ChR2(C128S)-EYFP ( Figure 5).

Discussion
Despite the fact that striatal MSNs have been strongly implicated in motor control, motivation and habit formation as well as neurological disorders like PD, Huntington's disease (HD), schizophrenia and addiction [35][36][37], a comprehensive understanding of their biological functions remains a major challenge. In addition to the classical anatomical studies [38,39], recent technical advances in neuroscience have started unveiling the intricate connections and functional significance of MSNs [34,40,41]. Nonetheless, a major obstacle in studying the functions of striatum is the presence of heterogeneous population of MSNs which are intermingled in the basal ganglia networks [42].
Although intervening striatal activity through optogenetic manipulation can be advantageous in above regards, a precise identification of activated neurons would be necessary to interpret the cellular and behavioral changes which are ensued. In mice, induction of c-fos mRNA or Fos immunoreactivity has been used as a proxy for photoactivation of neurons in different brain regions such as the hypothalamus [5], the cerebral cortex [4] and the striatum [40]. Unexpectedly, we did not observe c-fos induction after in vivo unilateral stimulation of the dorsal striatum ( Figure 2). The channelrhodopsin variant ChR2 was used in those previous studies and multiple light pulses were delivered over several minutes whereas we used the channelrhodopsin variant ChR2(C128S) and applied a single 500-ms pulse of blue light. ChR2 is photodepolarized rapidly and has a moderate channel closing rate while ChR2(C128S) has an increased light sensitivity with a slow off-kinetics [12,43,44]. Consequently, we were convinced that a number of experimental parameters such as the channelrhodopsin variant, number of light pulses and duration of stimulation could account for the absence of c-fos mRNA induction in our study. We also found that most of the commonly used IEGs were unsuitable to trace photoactivation in the mouse striatum. We then analyzed expression of Npas4 which was previously shown to be induced by in vivo sensory stimulation in the visual cortex [20] and has been implicated in learning and memory [45,46]. It has been reported that Npas4 is expressed in different brain regions, selectively in neurons and predominantly in excitatory neurons [19]. We found that in the striatum Npas4 expression was very low or absent at the basal condition ( Figure 2B2, Figure S1). Interestingly, in vivo optical illumination in freely moving mice differentially induced Npas4 expression specifically in the ipsilateral striatum ( Figure 2B2').
Can we also expect Npas4 induction following ChR2(C128S)mediated photostimulation in other neuronal cell types or other brain regions? Illumination to ChR2(C128S)-expressing hippocampal CA1 neurons resulted in induction of Npas4 [KFT unpublished data] in addition to c-fos induction [14]. However, induction of Npas4 was not apparent in raphe neurons where ChR2(C128S) was expressed by taking advantage of the tryptophan hydroxylase 2 (Tph2)-tTA line [KFT, unpublished data]. Thus, induction of a particular IEG, such as Npas4 and c-fos, following optical activation should be assessed for the neuronal cell type of interest.
Differential IEG induction following a given stimulus has also been demonstrated in previous studies [8,11,20]. Although expression of Npas4 was found to be selectively induced by membrane depolarization and Ca 2+ influx, Npas4 was not induced by several neurotrophic factors like BDNF and NT3 which readily induce other transcription factors such as c-fos and Arc [19,20]. It was found that after contextual fear conditioning (CFC) Npas4 induction took place as early as 5 minutes after training, much earlier than c-fos induction and the authors suggested that the pathways which induce Npas4 expression could be different from those for other IEGs [20]. Therefore, it is possible that in our study optical activation of ChR2(C128S)-expressing striatal MSNs triggered cellular pathways which were sufficient for induction of Npas4 but not for other IEGs like c-fos. Taken together, our results suggest that Npas4 can be a suitable tracer for identification of photoactivated MSNs at the cellular level in the mouse striatum considering the very low basal expression and a rapid and robust induction following stimulation.

Materials and Methods
Mice PDE10A2-tTA mice were obtained from RIKEN BRC Bank (RBRC No. RBRC02317, Strain B6.129-Pde10a2,tm1(tTA)Yok.) [13]. The tetO-ChR2(C128S)-EYFP BAC transgenic mice were crossed to PDE10A2-tTA mice to generate compound heterozygous mice [14]. Animals were kept under regulated air conditions (23uC61uC) and a 12:12 hours light-dark cycle throughout the experiments. Food and water were available ad libitum. Two to three mice were analyzed for each condition. All animal procedures were approved by the Animal Research Committees of the National Institute for Physiological Sciences, Keio University and Niigata University.

In vivo optical stimulation and electrophysiological recording of striatal neurons
To fix the head of the awake mouse in a stereotaxic apparatus, a small U-frame head holder was mounted on the head as reported previously [47]. Each mouse was anesthetized with ketamine hydrochloride (100 mg/kg body weight, i.p.) and xylazine hydrochloride (5 mg/kg body weight, i.p.) and fixed in a conventional stereotaxic apparatus (Narishige Scientific Instrument, Tokyo, Japan). The skull was widely exposed, and periosteum and blood on the skull were removed completely. The exposed skull was completely covered with bone adhesive resin (BISTITE II, Tokuyama, Tokyo, Japan) and acrylic resin (UNIFAST II, GC Corporation, Tokyo, Japan), and then a small U-frame head holder for head fixation was mounted and fixed with acrylic resin on the head of the mouse. After recovery from the first surgery (2 or 3 days later), the mouse was positioned in a stereotaxic apparatus with its head restrained using the U-frame head holder under light anesthesia with ketamine hydrochloride (50-100 mg/kg body weight, i.p.). A part of the skull in one hemisphere was removed to access the striatum.
After full recovery from the second surgery, the mouse was positioned in a stereotaxic apparatus with its head restrained using a U-frame head holder in the awake condition. For recording neural activity while illuminating with blue light, an electrode assembly consisting of a glass-coated Elgiloy microelectrode (0.5-1.0 MV at 1 kHz) and a 50 mm diameter optical fiber (Cer-amOptec Industries, East Longmeadow, MA, USA), was inserted perpendicularly into the brain through the dura mater using a hydraulic microdrive (Narishige Scientific Instrument, Tokyo, Japan). A blue laser (50 mW, CrystaLaser, Reno, NV, USA) was coupled to the optical fiber. The laser power was <40 mW at the fiber tip. The laser was controlled via TTL pulses driven by a stimulator (Nihon Kohden, Tokyo, Japan). The target area was 0.0-0.5 mm anterior and 2.0-2.2 mm lateral to bregma and 2.5-4.0 mm deep from the brain surface for the striatum [27]. Signals from the electrode were amplified, filtered (0.3-10 kHz), and sampled at 50 kHz using a computer.

In vivo optical stimulation in freely moving mice
Each mouse was anesthetized with ketamine hydrochloride and xylazine hydrochloride and fixed in a conventional stereotaxic apparatus (David Kopf, CA, USA). A plastic optical fiber (ESKA, Mitsubishi Rayon, Tokyo, Japan, 0.5 mm diameter) was inserted in each cerebral hemisphere above the dorsal striatum. The tip of the fiber located approximately at 0.7 mm posterior and 2.0 mm lateral to bregma, and 2.0 mm deep from the skull. The fiber was fixed on the skull using Aron Alpha (Toagosei Co., LTD., Tokyo, Japan). The mice were allowed to recover for at least one week after fiber implantation. For optical stimulation of striatal neurons, a single 500-ms illumination (6.7 mW/mm 2 at the fiber tip) was given to the left striatum in the home cage. No optical stimulation was given to the right striatum and this served as the sham-treated control. Mice were anesthetized with ketamine hydrochloride and xylazine hydrochloride after 5 minutes of optical stimulation and perfused transcardially after additional 5 minutes.

Double fluorescent in situ hybridization (Double FISH)
Procedures for double FISH were adopted from a previous study [14]. In brief, frozen tissue sections were hybridized with FITC-labeled Npas4 cRNA probe and DIG-labeled GFP, Chat or Pvalb cRNA probes. After stringent washing sections were incubated with a peroxidase-conjugated anti-FITC antibody (Roche, 1:200, 30 minutes at room temperature) and signals were visualized by FITC (TSA Plus Cyanine 3/Fluorescein System, PerkinElmer, Foster city, CA). Residual peroxidase activity was quenched by 2% H 2 O 2 (30 minutes at room temperature). Samples were then incubated with a peroxidase-conjugated anti-DIG antibody (Roche, 1:200, 60 minutes at room temperature) and signals were visualized by Cy3 (TSA Plus Cyanine 3/ Fluorescein System, PerkinElmer). Confocal fluorescence images were captured using a laser scanning microscope system (LSM 710, Carl Zeiss Microimaging, Germany). It is notable that in the standard chromogenic Npas4 ISH, very often we observed weaker perinuclear signals and stronger nucleolar signals (arrows, Figure  S3). In Double FISH using TSA amplification system, we mostly observed strong dot-like Npas4 signals.

Quantification of ISH signals
Images of stained coronal sections of the mouse brain were captured with an Olympus microscope (BX53, Olympus, Tokyo, Japan) and digital camera system (DP72, Olympus). The cells which were positive for ISH signals were counted in the dorsolateral striatum from five coronal tissue sections (approximately 60.5 mm bregma) from one representative animal. Student's t-test was performed to compare means. Difference between groups was considered highly significant when p#0.01 and significant when p#0.05.