Imaging epigenetic regulation by histone ...

NeuroImage 64 (2013) 630–639

Contents lists available at SciVerse ScienceDirect

NeuroImage journal homepage: www.elsevier.com/locate/ynimg

Imaging epigenetic regulation by histone deacetylases in the brain using PET/MRI with 18F-FAHA Hsin-Hsien Yeh a, 1, Mei Tian a, 1, Rainer Hinz b, 1, Daniel Young a, Alexander Shavrin a, Uday Mukhapadhyay a, Leo G. Flores a, Julius Balatoni a, Suren Soghomonyan a, Hwan J. Jeong a, Ashutosh Pal a, Rajesh Uthamanthil a, James N. Jackson a, Ryuichi Nishii c, Hiroshi Mizuma d, Hirotaka Onoe d, Shinya Kagawa e, Tatsuya Higashi e, Nobuyoshi Fukumitsu f, Mian Alauddin a, William Tong a, Karl Herholz b, Juri G. Gelovani a,⁎ a

Department of Experimental Diagnostic Imaging, University of Texas M.D. Anderson Cancer Center, Houston, Texas USA Wolfson Molecular Imaging Center, University of Manchester, Manchester, UK c Department of Radiology, Faculty of Medicine, University of Miyazaki, Miyazaki, Japan d Functional Probe Research Laboratory, RIKEN Center for Molecular Imaging Science, Kobe, Japan e Division of PET Imaging, Shiga Medical Center Research Institute, Moriyama, Japan f Department of Radiation Oncology, University of Tsukuba, Tsukuba, Japan b

a r t i c l e

i n f o

Article history: Accepted 5 September 2012 Available online 17 September 2012 Keywords: Epigenetics Histone deacetylases Positron emission tomography Brain Rhesus macaque

a b s t r a c t Epigenetic modifications mediated by histone deacetylases (HDACs) play important roles in the mechanisms of different neurologic diseases and HDAC inhibitors (HDACIs) have shown promise in therapy. However, pharmacodynamic profiles of many HDACIs in the brain remain largely unknown due to the lack of validated methods for noninvasive imaging of HDAC expression-activity. In this study, dynamic PET/CT imaging was performed in 4 rhesus macaques using [18F]FAHA, a novel HDAC substrate, and [18F]fluoroacetate, the major radio-metabolite of [18F]FAHA, and fused with corresponding MR images of the brain. Quantification of [18F]FAHA accumulation in the brain was performed using a customized dual-tracer pharmacokinetic model. Immunohistochemical analyses of brain tissue revealed the heterogeneity of expression of individual HDACs in different brain structures and cell types and confirmed that PET/CT/MRI with [ 18F]FAHA reflects the level of expression-activity of HDAC class IIa enzymes. Furthermore, PET/CT/MRI with [18F]FAHA enabled non-invasive, quantitative assessment of pharmacodynamics of HDAC inhibitor SAHA in the brain. © 2012 Elsevier Inc. All rights reserved.

Introduction The status of histone acetylation in the brain, that is regulated by cooperative functions of histone acetyl transferases (HATs) and histone deacetylases (HDACs), has received much attention during the last decade (Abel and Zukin, 2008; Jenuwein and Allis, 2001). Epigenetic mechanisms involving modifications of histones has been implicated in the development of various neurodegenerative diseases, multiple sclerosis, and epilepsy (Janssen et al., 2010; Kazantsev and Thompson, 2008; Mai et al., 2009; Monti et al., 2009; Tsankova et al., 2007). Histone deacetylase inhibitors (HDACIs) have shown efficacy in therapy of neurodegenerative diseases (Butler and Bates, 2006; Dietz and Casaccia, 2010; Sadri-Vakili and Cha, 2006). HDACIs increase the expression of

⁎ Corresponding author at: Department of Experimental Diagnostic Imaging, University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Avenue, room T8.3904, Houston, Texas 77030, USA. E-mail address: [email protected] (J.G. Gelovani). 1 These authors contributed equally. 1053-8119/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.neuroimage.2012.09.019

neuroprotective proteins, such as Hsp70 and Bcl-2 in the ischemic brain, which points to the therapeutic potential of HDACIs in stroke (Faraco et al., 2006). In models of motor neuron disease, such as the amyotrophic lateral sclerosis, inhibition of HDACs resulted in a reduction of clinical symptoms (Ryu et al., 2005). HDACIs have also shown promise in therapy of peripheral nerve damage (Cui et al., 2003). The involvement of HDACs in the mechanisms of development of alcoholism (Pandey et al., 2008) and drug addiction (Crepaldi and Riccio, 2009; Renthal and Nestler, 2009) and the effects of HDACIs in these disorders have been demonstrated as well (Kalda et al., 2007; Sanchis-Segura et al., 2009). However, the pharmacodynamic (PD) profiles of several HDACIs in the brain are largely unknown, which complicates the optimization and individualization of treatment protocols. The PD effects on HDACIs is often assessed in surrogate tissues (i.e., peripheral blood mononuclear cells Ronzoni et al., 2005) by measuring the levels of acetylated histones, which does not necessarily correlate with therapeutic responses in the brain and other tissues (Chavez-Blanco et al., 2005; Kelly et al., 2005). Also, it is not known whether there is a requirement for continuous

H.-H. Yeh et al. / NeuroImage 64 (2013) 630–639

inhibition of HDACs to achieve the optimal therapeutic effects and there are no predictive biomarkers for selection of patients who are most likely to respond to treatment with HDACIs (Kelly et al., 2005; Minucci and Pelicci, 2006). The assessment of pharmacodynamics of HDACIs in CNS is complicated due to an unacceptable traumatism of microdialysis and biopsy procedures. Therefore, novel molecular imaging agents for visualization and quantification of HDAC activity in the brain are urgently needed to facilitate the development and clinical translation of novel HDACIs. To address this need, we have been developing agents and methods for molecular imaging of pharmacodynamic effects of HDACIs using magnetic resonance spectroscopy (MRS) (Sankaranarayanapillai et al., 2006). Also, we developed a novel radiolabeled HDAC substrate, the 6-([ 18F]fluoroacetamido)-1-hexanoicanilide, termed [ 18F]FAHA, for non-invasive PET imaging of HDAC activity in the brain (Mukhopadhyay et al., 2006). Our initial in vitro and in vivo PET imaging studies in rats demonstrated a significant accumulation of [18F]FAHA in the brain already at 30 min after the radiotracer administration (Nishii et al., 2007). Subsequently, [18F]FAHA PET with simplified pharmacokinetic (PK) modeling was used to assess SAHA-induced inhibition of HDAC activity in the brain of non-human primates (Reid et al., 2009). In the current study, we performed PET/CT/MRI imaging with [ 18F] FAHA in rhesus macaques and quantified the rate of HDAC-mediated accumulation of [ 18F]FAHA in the brain using a new optimized pharmacokinetic model. To account for contribution of [ 18F]fluoroacetate ([ 18F]FACE), the major catabolite of [ 18F]FAHA, to the levels of [ 18F] FAHA-derived radioactivity measured in the brain, an additional PET/CT study with [18F]FACE was conducted in each animal. The results of in vivo PET/CTMRI studies were validated using radiochemical and immunohistochemical analyses of brain tissue, revealing the heterogeneity of expression of individual HDACs in different brain structures and cell types. This study demonstrated that PET/CT/MRI with [18F]FAHA reflects the level of expression-activity of HDAC class IIa enzymes in different structures of the brain and enables non-invasive, quantitative assessment of pharmacodynamics of HDAC inhibitor SAHA (vorinostat). Methods Radiosynthesis of [ 18F]FACE and [ 18F]FAHA The radiosynthesis of [18F]FACE was performed as described by us previously (Nishii et al., 2012). Specific activity of [18F]FACE was 1.162 Ci/μmol (43 GBq/μmol). The radiosynthesis of [18F]FAHA was performed as described by us previously (Mukhopadhyay et al., 2006). Decay-corrected yield was 11±2% and specific activity >2 GBq/mmol. Assessment of substrate specificity of FAHA for different HDACs In vitro enzyme assays were performed in a panel of classes I, II, III, and IV HDACs (BPS Bioscience, San Diego, CA) in the buffer: 25 mM Tris/Cl, pH 8.0, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, and 0.1 mg/ml BSA, 10 ng of each HDAC enzyme and different concentrations of non-radilabeled FAHA. Reactions were carried out for 30 min at 30 °C and products determined by HPLC using Ascentis RP-amide 7×300 mm column (Supelco), MeCN/0.02 mM O-phosphoric acid as mobile phase at 0.6 ml/min with UV detection of the cleaved product (hexanoic anilide) at 240 nm. The Km, Vmax and Kcat parameters were determined using software GraphPad Prism v4 (GraphPad Software, La Jolla, CA). Management of non-human primates All studies were performed under a protocol approved by the Institutional Animal Care and Use Committee of the University of Texas M.D. Anderson Cancer Center (UTMDACC). Four rhesus macaques (2 males and 2 females) under anesthesia were studied. Prior to imaging

631

studies, the animals were fasted overnight, but had a free access to water. The animals were premedicated with atropine sulfate 0.04 mg/kg intramuscularly (i.m.) and sedated with ketamine 10–15 mg/kg i.m., followed by intubation and inhalation anesthesia with isoflurane 1–3% in oxygen via Excel 210 SE (Ohmeda). Then, the animals were cannulated in both saphenous veins: one, for injection of 18F-FAHA or 18F-FACE; the other, for blood sampling. Body temperature was maintained at 37 °C with warming blanket (Model 505; Arizant Healthcare). Electrocardiogram, respiration, blood pressure and oxygenation were monitored with Solar 8000 (Marquette Medical Systems, WI). MRI/PET/CT imaging studies MR imaging was performed on an 8-channel 1.5 T HDx scanner with CRM gradient subsystem (GE Healthcare, WI). T1-weighted images were acquired using an inversion recovery-prepared 3D fast spoiled gradient recalled echo (FSPGR) sequence (12 degree flip angle, TE/TR=4.9/ 10.4 ms, 18×18 cm FOV, 320×320 acquisition matrix,1.2-mm sections reconstructed every 0.6 mm, ±21 kHz bandwidth). T2-weighted images were acquired using a 3D T2 CUBE imaging sequence (TE/TR= 128/3000 ms, echo train length: 120, 18×18 cm FOV, 320×320 acquisition matrix, 1.2-mm sections reconstructed every 0.6 mm, ±31 kHz bandwidth). The T1- and T2-weighted images were reformatted into 1-mm thick contiguous images in three orthogonal planes. PET/CT imaging was performed using Discovery STE16 (GE Healthcare, WI); the axial fields-of-view (FOV) of CT and PET were 50 cm and 70 cm, respectively. Transmission CT images of the head was used to verify the accuracy of positioning to match the prior MR images. The first PET imaging study was performed with 18F-FACE (5.95± 0.96 mCi in 5 ml saline) administered as a steady bolus over 1 min, as described before (Nishii et al., 2012; Tian et al., 2011). Dynamic PET images were acquired over 30 min. The second PET imaging study with [18F]FAHA (6.07±0.75 mCi/5 ml) was performed one week later using a similar protocol. Venous blood samples (0.5 ml) were obtained at: 0, 1, 2, 3, 5, 10 and 30 min for the analysis of radiolabeled metabolites ([18F]FACE and [18F]fluoride). PET images were reconstructed from the list-mode data for the following time frames: 12×15 s, 6×30 s, and 25×60 s; (total of 33 frames/study) using an iterative algorithm VuePoint (GE Healthcare) with 2 iterations, 28 subsets, 20 cm DFOV, kernel width 2.0 mm FWHM and a standard z-axial filter, with scatter and attenuation corrections. PET images had a resolution of 128×128 voxels with spatial resolution of 4.4 mm full width at half maximum (FWHM) and slice thickness of 3.25 mm. MR images were coregistered to the corresponding PET/CT images summed from dynamic 30 min scans using a rigid transformation algorithm and a normalized mutual method after running a re-slicing process with software PMOD version 3.1 (PMOD Technologies, Zurich, Switzerland). Volumes-of-interest (VOI) for carotid artery, whole brain, n. accumbens and cerebellum were defined and time activity curves (TACs) generated. Assessment of radiolabeled metabolites in blood plasma The radioactivity concentration in the whole blood and plasma was assayed using a gamma counter (Cobra, Packard, CT). Plasma was extracted from heparinized blood samples (0.5 ml) with 3× volumes of acetonitrile and analyzed with radio-HPLC: 1100 HPLC (Agilent, Santa Clara, CA), UV and radioactivity detectors (Bioscan Inc., Washington, DC); Supelcogel C-610H column (Sigma-Aldrich, MO); mobile phase: 35% MeCN/Phosphoric acid in water at 0.6 ml/min. This method affords separation of [ 18F]FAHA (elution at 18.1–18.4 min) and [ 18F]FACE (elution at 11.1–11.5 min). Fractions of [ 18F]FAHA and [ 18F]FACE were calculated for each sample based on the area under each peak. The total radioactivity in plasma was expressed as %ID/ml and plotted against time post injection of [ 18F]FAHA. The %ID/ml of [ 18F]FAHA and [ 18F]FACE were calculated based on the results of radio-HPLC analysis

632


and plotted against time. The average plasma-to-blood radioactivity concentration ratio was calculated from blood samples obtained at different times post 18F-FAHA administration and used for quantification of PET images, as described further. Assessment of [ 18F]FAHA-derived radiolabeled metabolites in the brain This study was performed under the approval of the RIKEN Kobe Institute Animal Committee (MAH21-13), Kobe, Japan. One rhesus macaque was anesthetized with ketamine (10 mg/kg) and atropine (0.08 mg/kg) for placement of an i.v. catheter. Then, the animal was placed into the specialized primate chair, and after recovery from anesthesia, [ 18F]FAHA (37 MBq/kg) was administered as a quick bolus injection. Three minutes later, the animal was deeply anesthetized by a bolus of pentobarbital (30 mg/kg, i.v.) and heparinized (1000U i.v.). At 7 min post injection of [ 18F]FAHA, the intracardiac perfusion with ice-cold 10 mM PBS was performed through the left ventricle and the animal sacrificed. The whole brain was quickly removed (within 10 min), sliced into 6 mm thick sections, and frozen in liquid nitrogen; different structures of the brain were dissected, weighed, and homogenized using an equal amount of ice-cold ddH2O. Brain tissue homogenates were mixed with 3 volumes of acetonitrile for 30 s and centrifuged at 10,000 ×g for 2 min at 4 °C. The supernatants were spotted (2 μl/spot) on a TLC plate silica 60 (Merck, Germany), developed using acetonitrile/water/ammonia hydroxide (90/9/1), and quantified by the imaging plate analyzer FLA-7000IR (Fujifilm, Japan). PET image analyses and pharmacokinetic modeling First, to calculate dynamic PET image-derived input functions for [ 18F]FAHA and [ 18F]FACE in blood plasma, the TAC was measured from the region of interest placed on the carotid artery and corrected using the mean plasma-to-whole blood radioactivity concentration ratio measured in blood samples (as described above). Then, fractional values of radioactivity concentrations of [ 18F]FAHA and [ 18F]FACE measured in venous blood plasma samples at different times post [ 18F]FAHA administration were applied to the image-derived plasma TAC to derive the corresponding TACs for [ 18F]FAHA and [ 18F]FACE, respectively. The newly developed pharmacokinetic model of [ 18F]FAHA-derived radioactivity accumulation in the brain, involves two blood plasma input functions for [ 18F]FAHA and [ 18F]FACE and three tissue compartments described in Fig. 1. In this model, we considered that [ 18F]FAHA can cross the blood–brain barrier (BBB) and cell membranes bi-directionally by non-facilitated diffusion, which is due to its physicochemical characteristics (LogD 1.44; PSA 58.2; FRD 8; LogBB − 0.51). Therefore, the rate constants of [ 18F]FAHA influx (k1FAHA) and efflux (k2FAHA) across the BBB and cell membranes were assumed to be high and not rate-limiting. Upon entry into the cell, [18F]FAHA is metabolized by HDACs to [18F]FACE and nonradiolabeled hexanoic anilyde, which is described by the rate constant k3FAHA. Because there are no known enzymes that would effectively catalyze the back-reaction to produce [18F]FAHA from [18F]FACE and hexanoic anilyde, the rate constant k4FAHA describing this reaction was set to zero. Also, we considered that a fraction of [18F]FACE produced inside the brain cells from [18F]FAHA can be trapped inside the cells due to high polarity and low lipophilicity of [18F]FACE, which limit its efflux (k2FACE) out of the cells. Furthermore, [18F]FACE can be progressively anabolized to [18F]fluoroacetyl-coA by acetyl-coA synthases ACS1 and ACS2, and by the citrate synthase to [18F]fluorocitrate, which binds irreversibly to aconitase (Fonnum et al., 1997; Proudfoot et al., 2006). However, there is no effective catabolism of fluoroacetyl-coA, fluorooxalate, and fluorocitrate back to fluoroacetate, which accounts for its neurotoxicity at pharmacologically relevant doses (Fonnum et al., 1997; Proudfoot et al., 2006). Therefore, the rate constant k4FACE was set to zero.

Fig. 1. Graphical representation of a novel multi-compartmental pharmacokinetic model with two simultaneous blood input functions for [18F]FAHA and [18F]FACE (see explanation in the Methods section).

Furthermore, we considered that [ 18F]FACE produced from catabolism of [ 18F]FAHA in organs and tissues other than brain, appears in plasma as the major radiolabeled metabolite and can contribute to radioactivity accumulation in the brain. It is known that [ 18F]FACE can cross BBB via the monocarboxylate transporter (MCT) and be selectively uptaken by the glial cells expressing MCT (Fonnum et al., 1997; Proudfoot et al., 2006). Therefore, the rate constants for influx (k1FACE) and efflux (k2FACE) of [ 18F]FACE across the BBB and cellular membranes were considered to be relatively smaller than those of [ 18F]FAHA. Importantly, we considered the rates of [ 18F]FACE efflux (k2FACE) and entrapment through anabolism to [ 18F]fluorocitrate (k3FACE) inside the cells to be independent of the origin of the [ 18F]FACE (whether it is produced from [ 18F]FAHA inside the cell or is transported into the cell from the outside, as the major metabolite of [ 18F]FAHA in blood). This approach requires a separate dynamic PET imaging study with [ 18F]FACE to calculate the rate constants k1FACE, k2FACE, and k3FACE for the brain. To achieve this, brain TACs of [ 18F]FACE-derived radioactivity were fitted to the irreversible two-tissue compartment model (Fig. 1) with a variable blood volume to estimate the rate constants k1FACE, k2FACE, k3FACE, using the dynamic imaging-derived [ 18F] FACE input function corrected by blood plasma sampling results, as described above in detail. Because there is no catabolism of fluorocitrate back into fluoroacetate, the k4FACE was set to zero and the unidirectional accumulation rate constant of [ 18F]FACE (KiFACE) was calculated as: KiFACE ¼ k1FACE ⋅k3FACE =ðk2FACE þ k3FACE Þ:

ð1Þ

Subsequently, the rate constants k1FACE, k2FACE, and k3FACE obtained from the PET imaging study with [ 18F]FACE alone were used as fixed rate constants in the model of [ 18F]FAHA-derived radioactivity accumulation in the brain (Fig. 1). The unidirectional accumulation rate constant of [18F]FAHA (KiFACE) was calculated as: KiFAHA ¼ k1FAHA ⋅k3FAHA =ðk2FAHA þ k2FACE Þ:

ð2Þ

Compartmental modeling, pharmacokinetic analyses, and generation of pixel-by-pixel parametric images were accomplished using PMOD 3.1 software (PMOD Technologies Ltd., Zurich, Switzerland). Statistical analyses of data were performed using GraphPad Prism 4 (GraphPad software, La Jolla, CA, USA).


633

Fig. 2. Enzymatic activity of different HDACs with FAHA. (A) Kcat data for different HDACs are shown in full scale and (B) the same data as in panel A are shown on a smaller scale to better visualize differences in Kcat data for HDACs classes I, II, III, and IV.

Assessment of SAHA-mediated dose-dependent inhibition of [ 18F]FAHA accumulation in the brain Four rhesus macaques (2 males and 2 females) were used to obtain the dose–response profiles of the impact of [ 18F]FAHA accumulation in the brain by vorinostat (SAHA). PET imaging studies were performed in each rhesus macaque: (a) baseline study with [ 18F]FACE; (b) baseline study with [ 18F]FAHA; and (c) 7 separate studies with [18F]FAHA after treatment with different doses of SAHA (0.001, 0.01, 0.1, 1, 10, 50, 100 mg/kg i.v.), which was administered as a continuous infusion over 1 h prior to [ 18F]FAHA administration. The time-course of radioactivity of [18F]FAHA in the brain after administration of escalating doses of selected HDAC inhibitor was assessed and compared to the baseline. Immunohistochemical staining of class I and II HDACs and acetyl-histones in the brain Formalin-fixed, paraffin-embedded brain samples were cut (5 μm), deparaffinized, rehydrated, and microwaved in 10 mM citrate buffer (pH 6.0) at 95 °C for 10 min; then incubated in 3% H2O2 for 15 min and blocked with 1.5% rabbit or goat serum in PBS for 60 min. Then, sections were incubated overnight at 4 °C with primary antibodies diluted 1:50 in 1% TBS: HDACs 1, 2, 3, 5, 7, 8, and 9 (Santa Cruz Biotech., CA);

HDACs 4 and 6 (Cell Signaling Technology, MA), HDACs 10–11 (Sigma-Aldrich, MO); and AH2A-K5, AH2B-K5, AH2B-K20, AH3-K9, AH3-K18, AH3-K23, AH4-K8, and AH4-K12 (Cell Signaling, MA). Biotinylated secondary antibodies and 3,3-diaminobenzidine from the Vectastain Elite kit (Vector Laboratories, CA) were used according to the manufacturer's protocol. The sections were counterstained with hematoxylin and evaluated on BX51 microscope, equipped with DP71 digital camera (Olympus, Japan). Results Substrate specificity of FAHA to class IIa HDAC enzymes In vitro enzymatic studies demonstrated that FAHA has higher substrate specificity for class IIa HDACs with Kcat values two orders of magnitude higher than for other classes of HDACs (Fig. 2). The time course of [ 18F]FACE and radiolabeled metabolites in blood and brain After intravenous administration of [ 18F]FACE, the clearance of radioactivity from plasma followed a bi-exponential kinetics with the fast phase between 3 and 18 min with a half-life of 4 min, followed

Fig. 3. Pharmacokinetics of [18F]FACE after i.v. administration. (A) Time-course of [18F]FACE (parent) and [18F]fluoride (metabolite) radioactivity in blood plasma, expressed as %ID/ml, and (B) as fraction of total radioactivity. (C) Comparison of [18F]FACE-derived time-activity curves in n. accumbens, cerebellum, and blood. Data — average±st.dev.

634


by a slow phase with a half-life of about 4 h (Fig. 3A). At 30 min post [ 18F]FACE administration, only negligible traces of [18F]fluoride were detectable in the blood plasma (b 2% of total radioactivity) (Fig. 3B). Partition of [ 18F]FACE-derived radioactivity in blood between RBCs and plasma was 70.2 ± 8.4%. Dynamic PET/CT/MRI imaging demonstrated very low, uniform distribution of [ 18F]FACE-derived radioactivity (Fig. 4A). The levels of [ 18F]FACE-derived radioactivity concentration between 0 and 30 min post injection were lower than that in blood plasma (Fig. 3C). The whole brain-to-blood and brain-to-muscle concentration ratios of [ 18F]FACE at 30 min post injection remained low, 0.55 ± 0.18 and 0.78 ± 0.28, respectively. Parametric images of [ 18F]FACE accumulation in the brain (Fig. 4B) also demonstrated very low Ki values in the forebrain and cerebellum, 0.006 ± 0.003 and 0.003 ± 0.003 min−1, respectively. The rate constants k1, k2, k3 and Ki of [ 18F]FACE in different structures of the brain are provided in Table 1.

The time course of [ 18F]FAHA and radiolabeled metabolites in blood and brain tissue After intravenous administration of [18F]FAHA, the clearance of total radioactivity from plasma followed a bi-exponential kinetics with the fast phase during the first 3 min with a half-life of 30 s followed by a slow phase with half-life of about 8 min (Fig. 5A). Systemic catabolism of [18F]FAHA to [18F]FACE was fast, which contributed to a rapid mono-exponential clearance of [18F]FAHA from the blood plasma with the fast phase during the first 2.5 min with half-life of 25 s followed by a slow phase with half-life of about 5.5 min. By 30 min post injection, almost all [18F]FAHA was catabolized to [18F]FACE (SI Appendix, Fig. S1). PET/CT/MRI imaging demonstrated a rapid accumulation of [ 18F] FAHA-derived radioactivity in the n. accumbens, amygdala, putamen, subthalamic nuclei, periaqueductal gray, inferior collicules, cerebellum,

Fig. 4. Axial PET/MRI fusion images of the brain obtained after i.v. administration of [18F]FACE or [18F]FAHA obtained at different axial levels. Static PET/MRI images were obtained at 30 min post administration of [18F]FACE (A) or [18F]FAHA (C) color-coded to the range of %ID/g values. Parametric PET/MRI images of the unidirectional accumulation rate constant Ki of [18F]FACE (B) or [18F]FAHA at baseline (D) and after escalating doses of SAHA (E–J) obtained in the same animal and color-coded to the same range of Ki values (min−1).


635

Fig. 5. Pharmacokinetics of [18F]FAHA after i.v. administration. Time-course of parent [18F]FAHA radioactivity at baseline and after pretreatment of animals with increasing doses of SAHA (color-coded in mg/kg) in (A) blood plasma, (B) n. accumbens, and (C) cerebellum. Data — average ± st.dev.

and especially in the vermis (Fig. 4C). Parametric images of unidirectional rate (Ki) of [ 18F]FAHA-derived radioactivity accumulation in the brain, generated using the newly developed compartmental model, enhanced these observations (Fig. 4D). The whole brain Ki of [18F]FAHA (0.145 ± 0.123 min−1) was about 50-fold higher than Ki of [ 18F]FACE (0.003 ± 0.002 min−1). Higher levels of [ 18F]FAHA-derived radioactivity accumulation in different nuclei of the brain and in the cerebellum were due to higher k3 rate constant values, reflecting higher activity of HDAC class IIa enzymes. The influx (k1) and efflux (k2) rate constants of [ 18F]FAHA in different structures of the brain were 10 to 20 fold higher than those of [ 18F]FACE, which is consistent with higher lipophilicity of [ 18F]FAHA. The rate constants k1, k2, k3, and Ki of [18F]FAHA for different structures of the brain are summarized in Table 1.

Radiolabeled metabolites in the brain Analyses of radiolabeled metabolites in brain tissue extracts obtained at 8–10 min post [ 18F]FAHA administration revealed [ 18F]

fluorocitrate as the major radiolabeled metabolite entrapped in the brain tissue (SI Appendix, Fig. S2).

Dose-dependent changes in [ 18F]FAHA pharmacokinetics, metabolism, and accumulation in different structures of the brain Pretreatment of animals with increasing doses of SAHA resulted in a dose-dependent log-linear increase in circulation half-time and AUC of [18F]FAHA and corresponding decrease in AUC of [18F]FACE radiometabolite (SI Appendix, Fig. S1). In contrast, a dose-dependent inhibition of accumulation and the rate constant (Ki) of [ 18F]FAHA were observed in different structures of the brain with increasing doses of SAHA (Fig. 4E–J). The observed dose-dependent inhibition of [ 18F]FAHA accumulation was due to reduction of the rate constant k3 for conversion of [ 18F]FAHA into [ 18F]FACE (Table 1). The IC50 values of SAHA were calculated using Ki [ 18F]FAHA values for different structures of the brain (Fig. 6; Table 2). Higher IC50 values for SAHA were observed in periaqueductal gray, cerebellum, n. accumbens, inferior

Fig. 6. Dose dependent inhibition of the [18F]FAHA accumulation in different structures of the brain. SAHA dose–response curves based on [18F]FAHA rate constants k3 and Ki are shown for representative structures of the brain. Data — average ± st.err.

636


Table 1 The rate constants influencing the accumulation of [18F]FACE and [18F]FAHA in different structures of the brain. Brain regions

Whole brain Putamen Thalamus Nucleus accumbens Temporal cortex Amygdala Peri-aqueductal gray Inferior colliculus Hippocampus Cerebellum White matter

FACE baseline

FAHA baseline

k1

k2

k3

Ki

k1

k2

k3

Ki

0.045 ± 0.034 0.024 ± 0.033 0.020 ± 0.482 0.056 ± 0.052 0.016 ± 0.046 0.020 ± 0.035 0.021 ± 0.044 0.019 ± 0.045 0.021 ± 0.034 0.055 ± 0.054 0.015 ± 0.019

0.170 ± 0.145 0.059 ± 0.063 0.048 ± 0.067 0.179 ± 0.146 0.037 ± 0.014 0.088 ± 0.063 0.050 ± 0.034 0.034 ± 0.024 0.118 ± 0.031 0.202 ± 0.125 0.089 ± 0.103

0.033 ± 0.065 0.026 ± 0.031 0.022 ± 0.022 0.025 ± 0.034 0.033 ± 0.018 0.047 ± 0.040 0.026 ± 0.023 0.022 ± 0.021 0.054 ± 0.039 0.044 ± 0.066 0.029 ± 0.022

0.003 ± 0.002 0.008 ± 0.003 0.008 ± 0.003 0.004 ± 0.003 0.008 ± 0.004 0.006 ± 0.004 0.007 ± 0.003 0.008 ± 0.004 0.006 ± 0.003 0.003 ± 0.003 0.004 ± 0.002

0.424 ± 0.2637 0.708 ± 0.3726 0.765 ± 0.4308 0.626 ± 0.5812 0.438 ± 0.2453 0.512 ± 0.1283 0.619 ± 0.2961 0.706 ± 0.2366 0.595 ± 0.071 0.645 ± 0.2594 0.463 ± 0.2052

0.611 ± 0.5713 0.731 ± 0.617 0.604 ± 0.365 1.181 ± 1.6532 0.558 ± 0.2452 0.493 ± 0.118 0.554 ± 0.256 0.644 ± 0.312 0.883 ± 0.738 1.001 ± 0.475 0.810 ± 0.269

0.192 ± 0.222 0.328 ± 0.079 0.192 ± 0.032 0.401 ± 0.296 0.158 ± 0.056 0.259 ± 0.093 0.236 ± 0.088 0.267 ± 0.108 0.279 ± 0.213 0.380 ± 0.279 0.187 ± 0.135

0.145 ± 0.123 0.295 ± 0.367 0.223 ± 0.165 0.663 ± 0.913 0.116 ± 0.191 0.242 ± 0.158 0.260 ± 0.080 0.301 ± 0.234 0.240 ± 0.155 0.272 ± 0.024 0.111 ± 0.012

colliculus, putamen, and amygdala. Relatively lower SAHA IC50 values were observed in the hippocampus, temporal cortex, and white matter. Expression of class I and II HDACs and acetylation status of histones in different brain structures In the n. accumbens (SI Appendix, Fig. S3), HDACs class I and class IIa, were highly expressed, predominantly in the neurons, except for HDAC 7, which was expressed in microvessels. In particular, high levels of HDACs 4 and 5 expression were observed in neurons. HDAC 6 was expressed in neurons and glia, whereas HDAC 10 was expressed predominantly in glial cells. HDAC 11 was highly expressed both, in neurons and glia (Fig. S4). H3-K18 was most deacetylated, while H2B-K20 and H4K-12 were most acetylated. In the hippocampus (SI Appendix, Figs. S5–9), HDACs 2, 3, and 10 were expressed at high levels, especially in the dentate gyrus (DG) neurons. HDAC 5 was moderately expressed in CA2 and CA3 neurons, whereas HDACs 1 and 4 were expressed the least (SI Appendix, Figs. S6, S7). HDAC 6 was moderately expressed in CA1 and DG (SI Appendix, Figs. S5, S9). Moderate-to-high levels of HDAC 11 were observed throughout hippocampus (SI Appendix, Fig. S5). H3-K18 was the most acetylated, while H3-K9 and H3-K23 were the most deacetylated throughout the hippocampus (SI Appendix, Fig. S11–14). In the brain cortex (SI Appendix, Fig. S15), the level of expression of class IIa HDACs was notably lower than that of the class I HDACs, and lower than the expression of HDAC class IIa in the n. accumbens. HDAC 6 was expressed in the neurons and glia, whereas HDAC 10 was expressed predominantly in the glia. HDAC 11 was expressed both, in neurons and glia. H3-K18 was the most acetylated, while H3-K9 and H3-K23 were the most deacetylated (SI Appendix, Fig. S16). In the cerebellum (SI Appendix, Fig. S17), amongst class I HDACs, high levels of HDACs 1 and 2 were observed in Bergmann microglial cells, whereas high levels of 8 expression were observed in most of the cell types. Amongst HDAC class IIa, high levels of HDACs 4 and 9

Table 2 IC50 values of SAHA calculated using Ki [18F]FAHA values for different structures of the brain. Brain Structure

IC50

Std.Err.

Whole brain Putamen Thalamus N. accumbens Temporal cortex Amygdala Peri-aquedutal gray Inferior colliculus Hippocampus Cerebellum White matter

6.94 3.34 1.81 4.47 1.16 3.21 8.28 4.42 0.10 6.10 0.32

0.53 0.46 0.51 0.60 0.52 0.39 0.37 0.36 0.93 0.45 0.77

expression was observed in Purkinje cells, whereas HDAC 5 was moderately expressed in granular and molecular layer cells. HDAC 7 was expressed in microvascular endothelial cells and as granules localized in the pinceau terminals of basket neurons, embracing the Purkinje cell bodies. Amongst HDACs class IIb, HDAC 6 was expressed predominantly in the Purkinje cells, whereas HDAC 10 was expressed in granular and molecular layer cells. HDAC 11 (class IV) was moderately expressed in Brenner microglial cells. The acetylation of H2A-K5, H2B-K20, H3-K9, H3-K18, H3-K23, and H4-K12, was almost non-detectable in the vast majority of Purkinge cells. Low acetylation levels of H3-K23 were observed in all cells in the cerebellum, whereas the lowest levels of H3-K9 acetylation was observed in the granular layer cells. Highest levels of acetylation of other histone lysine residues were observed in cells of the molecular layer (SI Appendix, Fig. S18). The results of qualitative assessment of expression of individual HDAC isoforms in different brain structures are summarized in SI Appendix, Table S1. The acetylation status of certain lysine residues on different histones is summarized in SI Appendix, Table S2. Discussion There is increasing evidence that HDAC-mediated epigenetic regulation is involved in neurogenesis, neuronal plasticity, learning, memory formation and recall, as well as CNS disorders including cognitive dysfunction, depression, addiction, and schizophrenia and several neurodegenerative diseases (for review: Fischer et al., 2010; Jiang et al., 2008; Renthal and Nestler, 2009; Tsankova et al., 2007). Development of novel isotype-selective HDAC inhibitors and the assessment of their efficacy for therapy of different brain diseases (Thomas, 2009) could be facilitated by PET imaging with HDAC class and isotype-selective agents. In this study, we determined that [ 18F]FAHA exhibits preferential substrate affinity to class IIa enzymes (HDACs 4, 5, 7, and 9), which is consistent with other reports demonstrating that trifluoro substitutions in the methyl group of acetyl moiety of acetyl-lysine, increase substrate specificity for the class IIa HDACs (Jones et al., 2008; Lahm et al., 2007). To quantify PET/CT/MR images of [ 18F]FAHA accumulation in the brain we developed a new pharmacokinetic compartmental model that is substantially different than previously reported simple oneand two-compartmental models (Reid et al., 2009). We demonstrated that due to high lipophilicity, [18F]FAHA crosses the cell membranes and BBB by non-facilitated diffusion, as evidenced by high influx and efflux rate constants k1FAHA and k2FAHA, respectively. In contrast, the influx and efflux rate constants of [ 18F]FACE are 10 to 20 fold smaller, because fluoroacetate is a hydrophilic molecule and does not cross membranes easily and its transport across the blood–brain-barrier and across glial and neuronal cell membranes is rate-limited by H+ coupled monocarboxylate transporters (MCTs). The MCT1 is expressed predominantly by the brain microvascular endothelial and glial cells,


whereas the MCT2 is expressed predominantly in neurons, especially in the cortex, hippocampus, and cerebellum (Pierre and Pellerin, 2005). Although after intravenous administration [ 18F]FAHA is gradually catabolized to [ 18F]FACE, the contribution of [ 18F]FACE circulating in the blood to radioactivity accumulation in the brain, especially during the first 30 min post injection, is very small, as evidenced by an almost 50-fold lower KiFACE, as compared to KiFAHA. Nevertheless, the contribution of [ 18F]FACE as the major metabolite of [ 18F]FAHA circulating in blood to accumulation of [ 18F]FAHA-derived radioactivity in the brain was accounted for in the new pharmacokinetic model. Inside the cells, [ 18F]FAHA is cleaved predominantly by the class IIa HDAC enzymes releasing [ 18F]FACE, which effluxes significantly slower than [ 18F]FAHA and causes retention of radioactivity in the brain. In the current study, by analyzing radiolabeled metabolites in the brain extracts we demonstrated that [ 18F]FACE produced from [ 18F]FAHA by enzymatic activity of HDACs is further anabolized to [18F]fluorocitrate. Fluorocitrate is known to irreversibly bind to aconitase (Carrell et al., 1970; Lauble et al., 1996) and is trapped inside the cells. The metabolic retention of [ 18F]fluorocitrate inside the cells is also evidenced by results of our previous studies on pharmacokinetics, metabolism, and radiation dosimetry of [ 18F]FACE, in which the [ 18F]fluorocitrate anabolite was not detectable in blood up to 3 h post radiotracer administration (Nishii et al., 2012). However, the rate of irreversible entrapment of [ 18F]FACE due to anabolism into [ 18F]fluorocitrate in the brain is slow, as evidenced by a relatively low k3FACE, when [ 18F]FACE was administered prior to [ 18F]FAHA imaging studies. This is consistent with the known biochemistry and toxicology of FACE; fluorocitrate formation from FACE in the brain is a relatively slow process, which results in a delayed neurotoxicity (Clarke, 1991; Goncharov et al., 2006; Peters and Wakelin, 1953). There are no known enzymes in the brain that would directly de-fluorinate [ 18F]FAHA or [ 18F]FACE, and which would catalyze reverse reactions from [ 18F]fluorocitrate back to [ 18F]FACE or [ 18F]FACE back to [18F]FAHA. Therefore, in the compartmental model used in this study, the rate constants k4FAHA and k4FACE were set to zero. An important observation made in this study was that the rate of HDAC-mediated cleavage of [18F]FAHA to [ 18F]FACE in the brain, defined as k3FAHA, was significantly faster than the rate of anabolism and irreversible entrapment of [ 18F]FACE defined as k3FACE. On the other hand, the efflux of [ 18F]FACE from the cells characterized by the rate constant k2FACE was 4 to 10 fold less than that of [ 18F]FAHA. Therefore, the accumulation of [ 18F]FAHA-derived radioactivity in different structures of the brain within the first 30 min post i.v. injection is ratelimited predominantly by the level of expression and activity of HDAC class IIa enzymes in corresponding structures. However, the observed pattern and magnitude of accumulation of [ 18F]FAHA-derived radioactivity in different structures of the brain reflect the activity of class IIa HDACs under general anesthesia, which may be different from that under non-anesthetized conditions. We decided to develop a radiolabeled substrate as opposed to radiolabeled inhibitor of HDACs, because binding of a radiolabeled inhibitor to its target enzyme reflects only the magnitude of enzyme expression/occupancy, but not necessarily the enzyme activity. Other investigators have explored the feasibility of PET imaging of HDAC expression using radiolabeled HDAC inhibitors, such as [18F]suberoylanilide hydroxamic acid, [18F]SAHA (Hendricks, et al., 2011; Zeglis, et al., 2011) and [11C]MS-275 (Hooker, et al., 2010). However, both [18F]SAHA and [11C]MS-275 did not accumulate in the brain due to either excessive washout, such as in the case of [18F]SAHA, or poor blood–brain barrier (BBB) permeability, such as in the case of [11C]MS-275. As for the site of radiolabeling, we elected not to label FAHA in sites other than in the fluoroacetyl moiety, because 6-amino-1-hexanoicanilide moiety produced as the result of HDAC-mediated cleavage of [18F]FAHA is lipophylic and cannot be entrapped inside the cell, as it freely diffuses across the cell membrane. The results of our immunohistochemical (IHC) demonstrated a link between the level of class IIa HDAC enzymes and the acetylation status

637

of certain lysine residues of several histone proteins in different structures of the brain, with the magnitude of [18F]FAHA accumulation in the corresponding structures. In particular, high levels of [18F]FAHA accumulation observed in n. accumbens, and cerebellum were associated with high expression levels of HDACs 4 and/or 5. These observations are important, because HDACs 4 and 5 in n. accumbens are known to play important roles in the development of drug addiction (Kumar et al., 2005; Renthal et al., 2007). Consistent with previous reports (Broide et al., 2007), in this study moderate levels of HDAC 5 expression and hypoacetylation of H3 and H4 were observed in the hippocampus, especially in CA2 and CA3 neurons, but lower levels in neurons of the dentate gyrus. HDAC 5 expression in hippocampal neurons plays an important role in the mechanism of depression (Tsankova et al., 2006). A stimulation inducing long-term depression causes HDAC 5-mediated deacetylation of H3 and H4 in the C/EBP gene promoter and prevents the establishment of long-term potentiation (Guan et al., 2002). Therefore, PET/CT/MRI with [ 18F]FAHA can expand the arsenal of neuroimaging approaches and facilitate the longitudinal studies of epigenetic mechanisms of addiction and depression. Higher expression levels of HDACs 4, 5, and 9 observed in the cerebellum explain higher levels of [ 18F]FAHA accumulation. Expression of HDAC 4 in Purkinje cells is consistent with survival promoting functions of HDAC 4 in Purkinje cells, which involves inhibition of cyclin-dependent kinase-1 (CDK1) and prevents abortive cell-cycle progression. In contrast, mice lacking HDAC 4 have elevated CDK1 activity and display cerebellar abnormalities including a progressive degeneration and loss of Purkinje neurons postnatally in posterior lobes (Majdzadeh et al., 2008). Expression of HDAC 5 observed in cerebellar granule cells is also consistent with its neuroprotective functions (Bolger and Yao, 2005; Chawla et al., 2003; Linseman et al., 2003). Noteworthy, is an almost complete deacetylation of lysine residues of histone proteins (H2A-K5, H2B-K20, H3-K9, H3-K18, H3-K23, and H4-K12) observed in Purkinje cells in the cerebellum, which warrants further studies to elucidate the significance of these new findings. In contrast to n. accumbens, hippocampus, and cerebellum, much lower levels of class IIa HDAC expression and lower activity and a more pronounced acetylation of histones were observed in the brain cortex and white matter, which explains the lower accumulation of [ 18F]FAHA in these tissues. Pre-treatment of animals with increasing doses of HDAC inhibitor SAHA (vorinostat) resulted in a dose-dependent decrease of the rate constants k3FAHA and KiFAHA, in all structures of the brain, despite an increase in [18F]FAHA blood input function (due to a decrease in systemic catabolism of [18F]FAHA to [18F]FACE). Consistent with the results of IHC studies, lower IC50 values for SAHA were determined for brain regions expressing lower levels of class IIa HDACs (i.e., cortex), as compared to higher IC50 values of SAHA for the brain structures expressing class IIa HDACs at higher levels (i.e., n. accumbens, cerebellum). Conclusions Taken together, these results indicate that the unidirectional accumulation of [18F]FAHA-derived radioactivity in the brain (KiFAHA) is rate-limited by the level of expression-activity of HDAC class IIa enzymes in various structures of the brain, defined in the current pharmacokinetic model by the rate constant k3FAHA. Therefore, PET imaging with [ 18F]FAHA could be used as a pharmacodynamic biomarker of inhibition of class IIa HDACs in the brain and facilitate the development and clinical translation of novel class-IIa HDACIs. Author contributions J.G. and W.T. developed the approach to imaging the expression/ activity of HDACs. U.M., J.B., A.P., and M.A. synthesized 18F-FACE and 18F-FAHA and performed the analysis of radiolabeled metabolites in blood.

638


A.S. conducted enzyme assays with FAHA and different HDACs. H.H.Y., M.T., L.G.F., S.S., H.J.J., N.F., R.U., and J.N.J. performed PET/CT and MR imaging and blood sampling studies in primates. R.N., H.M., H.O., S.K., and T.H. performed studies on radiolabeled metabolites in brain. H.H.Y., R.H., M.T., K.H., and J.G. optimized the pharmacokinetic model, performed image analysis and quantification, generated parametric PET images and PET/MRI fusion images. D.Y. performed all histopathology, immunohistochemical staining and interpreted results. J.G., H.H.Y., R.H., M.T., M.A., and K.H. wrote this paper. Acknowledgments This work was supported by the grant 1RC2DA028912 (NIDA, NIH) to JG. Appendix A. Supplementary data Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.neuroimage.2012.09.019. References Abel, T., Zukin, R.S., 2008. Epigenetic targets of HDAC inhibition in neurodegenerative and psychiatric disorders. Curr. Opin. Pharmacol. 8, 57–64. Bolger, T.A., Yao, T.P., 2005. Intracellular trafficking of histone deacetylase 4 regulates neuronal cell death. J. Neurosci. 25, 9544–9553. Broide, R.S., Redwine, J.M., Aftahi, N., Young, W., Bloom, F.E., Winrow, C.J., 2007. Distribution of histone deacetylases 1–11 in the rat brain. J. Mol. Neurosci. 31, 47–58. Butler, R., Bates, G.P., 2006. Histone deacetylase inhibitors as therapeutics for polyglutamine disorders. Nat. Rev. Neurosci. 7, 784–796. Carrell, H.L., Glusker, J.P., Villafranca, J.J., Mildvan, A.S., Dummel, R.J., Kun, E., 1970. Fluorocitrate inhibition of aconitase: relative configuration of inhibitory isomer by x-ray crystallography. Science 170, 1412–1414. Chavez-Blanco, A., Segura-Pacheco, B., Perez-Cardenas, E., Taja-Chayeb, L., Cetina, L., Candelaria, M., Cantu, D., Gonzalez-Fierro, A., Garcia-Lopez, P., Zambrano, P., PerezPlasencia, C., Cabrera, G., Trejo-Becerril, C., Angeles, E., Duenas-Gonzalez, A., 2005. Histone acetylation and histone deacetylase activity of magnesium valproate in tumor and peripheral blood of patients with cervical cancer. A phase I study. Mol. Cancer 4, 22. Chawla, S., Vanhoutte, P., Arnold, F.J., Huang, C.L., Bading, H., 2003. Neuronal activitydependent nucleocytoplasmic shuttling of HDAC4 and HDAC5. J. Neurochem. 85, 151–159. Clarke, D.D., 1991. Fluoroacetate and fluorocitrate: mechanism of action. Neurochem. Res. 16, 1055–1058. Crepaldi, L., Riccio, A., 2009. Chromatin learns to behave. Epigenetics 4, 23–26. Cui, S.S., Yang, C.P., Bowen, R.C., Bai, O., Li, X.M., Jiang, W., Zhang, X., 2003. Valproic acid enhances axonal regeneration and recovery of motor function after sciatic nerve axotomy in adult rats. Brain Res. 975, 229–236. Dietz, K.C., Casaccia, P., 2010. HDAC inhibitors and neurodegeneration: at the edge between protection and damage. Pharmacol. Res. 62, 11–17. Faraco, G., Pancani, T., Formentini, L., Mascagni, P., Fossati, G., Leoni, F., Moroni, F., Chiarugi, A., 2006. Pharmacological inhibition of histone deacetylases by suberoylanilide hydroxamic acid specifically alters gene expression and reduces ischemic injury in the mouse brain. Mol. Pharmacol. 70, 1876–1884. Fischer, A., Sananbenesi, F., Mungenast, A., Tsai, L.H., 2010. Targeting the correct HDAC(s) to treat cognitive disorders. Trends Pharmacol. Sci. 31, 605–617. Fonnum, F., Johnsen, A., Hassel, B., 1997. Use of fluorocitrate and fluoroacetate in the study of brain metabolism. Glia 21, 106–113. Goncharov, N.V., Jenkins, R.O., Radilov, A.S., 2006. Toxicology of fluoroacetate: a review, with possible directions for therapy research. J. Appl. Toxicol. 26, 148–161. Guan, Z., Giustetto, M., Lomvardas, S., Kim, J.H., Miniaci, M.C., Schwartz, J.H., Thanos, D., Kandel, E.R., 2002. Integration of long-term-memory-related synaptic plasticity involves bidirectional regulation of gene expression and chromatin structure. Cell 111, 483–493. Hendricks, J.A., Keliher, E.J., Marinelli, B., Reiner, T., Weissleder, R., Mazitschek, R., 2011. In vivo PET imaging of histone deacetylases by 18F-suberoylanilide hydroxamic acid (18F-SAHA). J. Med. Chem. 54, 5576–5582. Hooker, J.M., Kim, S.W., Alexoff, D., Xu, Y., Shea, C., Reid, A., Volkow, N., Fowler, J.S., 2010. Histone deacetylase inhibitor, MS-275, exhibits poor brain penetration: PK studies of [C]MS-275 using Positron Emission Tomography. ACS Chem. Neurosci. 1, 65–73. Janssen, C., Schmalbach, S., Boeselt, S., Sarlette, A., Dengler, R., Petri, S., 2010. Differential histone deacetylase mRNA expression patterns in amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 69, 573–581. Jenuwein, T., Allis, C.D., 2001. Translating the histone code. Science 293, 1074–1080.

Jiang, Y., Langley, B., Lubin, F.D., Renthal, W., Wood, M.A., Yasui, D.H., Kumar, A., Nestler, E.J., Akbarian, S., Beckel-Mitchener, A.C., 2008. Epigenetics in the nervous system. J. Neurosci. 28, 11753–11759. Jones, P., Altamura, S., De Francesco, R., Gallinari, P., Lahm, A., Neddermann, P., Rowley, M., Serafini, S., Steinkuhler, C., 2008. Probing the elusive catalytic activity of vertebrate class IIa histone deacetylases. Bioorg. Med. Chem. Lett. 18, 1814–1819. Kalda, A., Heidmets, L.T., Shen, H.Y., Zharkovsky, A., Chen, J.F., 2007. Histone deacetylase inhibitors modulates the induction and expression of amphetamineinduced behavioral sensitization partially through an associated learning of the environment in mice. Behav. Brain Res. 181, 76–84. Kazantsev, A.G., Thompson, L.M., 2008. Therapeutic application of histone deacetylase inhibitors for central nervous system disorders. Nat. Rev. Drug Discov. 7, 854–868. Kelly, W.K., O'Connor, O.A., Krug, L.M., Chiao, J.H., Heaney, M., Curley, T., MacGregoreCortelli, B., Tong, W., Secrist, J.P., Schwartz, L., Richardson, S., Chu, E., Olgac, S., Marks, P.A., Scher, H., Richon, V.M., 2005. Phase I study of an oral histone deacetylase inhibitor, suberoylanilide hydroxamic acid, in patients with advanced cancer. J. Clin. Oncol. 23, 3923–3931. Kumar, A., Choi, K.H., Renthal, W., Tsankova, N.M., Theobald, D.E., Truong, H.T., Russo, S.J., Laplant, Q., Sasaki, T.S., Whistler, K.N., Neve, R.L., Self, D.W., Nestler, E.J., 2005. Chromatin remodeling is a key mechanism underlying cocaine-induced plasticity in striatum. Neuron 48, 303–314. Lahm, A., Paolini, C., Pallaoro, M., Nardi, M.C., Jones, P., Neddermann, P., Sambucini, S., Bottomley, M.J., Lo Surdo, P., Carfi, A., Koch, U., De Francesco, R., Steinkuhler, C., Gallinari, P., 2007. Unraveling the hidden catalytic activity of vertebrate class IIa histone deacetylases. Proc. Natl. Acad. Sci. U. S. A. 104, 17335–17340. Lauble, H., Kennedy, M.C., Emptage, M.H., Beinert, H., Stout, C.D., 1996. The reaction of fluorocitrate with aconitase and the crystal structure of the enzyme-inhibitor complex. Proc. Natl. Acad. Sci. U. S. A. 93, 13699–13703. Linseman, D.A., Bartley, C.M., Le, S.S., Laessig, T.A., Bouchard, R.J., Meintzer, M.K., Li, M., Heidenreich, K.A., 2003. Inactivation of the myocyte enhancer factor-2 repressor histone deacetylase-5 by endogenous Ca(2+)//calmodulin-dependent kinase II promotes depolarization-mediated cerebellar granule neuron survival. J. Biol. Chem. 278, 41472–41481. Mai, A., Rotili, D., Valente, S., Kazantsev, A.G., 2009. Histone deacetylase inhibitors and neurodegenerative disorders: holding the promise. Curr. Pharm. Des. 15, 3940–3957. Majdzadeh, N., Morrison, B.E., D'Mello, S.R., 2008. Class IIA HDACs in the regulation of neurodegeneration. Front. Biosci. 13, 1072–1082. Minucci, S., Pelicci, P.G., 2006. Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat. Rev. Cancer 6, 38–51. Monti, B., Polazzi, E., Contestabile, A., 2009. Biochemical, molecular and epigenetic mechanisms of valproic acid neuroprotection. Curr. Mol. Pharmacol. 2, 95–109. Mukhopadhyay, U.T.W., Gelovani, J., Alauddin, M.M., 2006. Radiosynthesis of 6( 18F-fluoroacetamido)-1-hexanoicanilide ( 18F-FAHA) for PET Imaging of histone deacetylase (HDAC). J. Labelled Comp. Radiopharm. 49, 997–1006. Nishii, R.M.U., Yeh, H.-H., Soghomonyan, S., Volgin, A., Alauddin, M.M., Tong, W., Gelovani, J.G., 2007. PET imaging of histone deacetylase activity in a rat brain using 6-([18F]-fluoroacetamide)-1-hexanoicanilide ([18F]-FAHA). J. Nucl. Med. 48 (Supplement 2) (336 pp.). Nishii, R., Tong, W., Wendt III, R., Soghomonyan, S., Mukhopadhyay, U., Balatoni, J., Mawlawi, O., Bidaut, L., Tinkey, P., Borne, A., Alauddin, M., Gonzalez-Lepera, C., Yang, B., Gelovani, J.G., 2012. Pharmacokinetics, metabolism, biodistribution, radiation dosimetry, and toxicology of (18)F-fluoroacetate ((18)F-FACE) in non-human primates. Mol. Imaging Biol. 14 (2), 213–224. Pandey, S.C., Ugale, R., Zhang, H., Tang, L., Prakash, A., 2008. Brain chromatin remodeling: a novel mechanism of alcoholism. J. Neurosci. 28, 3729–3737. Peters, R., Wakelin, R.W., 1953. Biochemistry of fluoroacetate poisoning; the isolation and some properties of the fluorotricarboxylic acid inhibitor of citrate metabolism. Proc. R. Soc. Lond. B Biol. Sci. 140, 497–507. Pierre, K., Pellerin, L., 2005. Monocarboxylate transporters in the central nervous system: distribution, regulation and function. J. Neurochem. 94, 1–14. Proudfoot, A.T., Bradberry, S.M., Vale, J.A., 2006. Sodium fluoroacetate poisoning. Toxicol. Rev. 25, 213–219. Reid, A.E., Hooker, J., Shumay, E., Logan, J., Shea, C., Kim, S.W., Collins, S., Xu, Y., Volkow, N., Fowler, J.S., 2009. Evaluation of 6-([(18)F]fluoroacetamido)-1-hexanoicanilide for PET imaging of histone deacetylase in the baboon brain. Nucl. Med. Biol. 36, 247–258. Renthal, W., Nestler, E.J., 2009. Histone acetylation in drug addiction. Semin. Cell Dev. Biol. 20, 387–394. Renthal, W., Maze, I., Krishnan, V., Covington III, H.E., Xiao, G., Kumar, A., Russo, S.J., Graham, A., Tsankova, N., Kippin, T.E., Kerstetter, K.A., Neve, R.L., Haggarty, S.J., McKinsey, T.A., Bassel-Duby, R., Olson, E.N., Nestler, E.J., 2007. Histone deacetylase 5 epigenetically controls behavioral adaptations to chronic emotional stimuli. Neuron 56, 517–529. Ronzoni, S., Faretta, M., Ballarini, M., Pelicci, P., Minucci, S., 2005. New method to detect histone acetylation levels by flow cytometry. Cytometry A 66, 52–61. Ryu, H., Smith, K., Camelo, S.I., Carreras, I., Lee, J., Iglesias, A.H., Dangond, F., Cormier, K.A., Cudkowicz, M.E., Brown Jr., R.H., Ferrante, R.J., 2005. Sodium phenylbutyrate prolongs survival and regulates expression of anti-apoptotic genes in transgenic amyotrophic lateral sclerosis mice. J. Neurochem. 93, 1087–1098. Sadri-Vakili, G., Cha, J.H., 2006. Histone deacetylase inhibitors: a novel therapeutic approach to Huntington's disease (complex mechanism of neuronal death). Curr. Alzheimer Res. 3, 403–408. Sanchis-Segura, C., Lopez-Atalaya, J.P., Barco, A., 2009. Selective boosting of transcriptional and behavioral responses to drugs of abuse by histone deacetylase inhibition. Neuropsychopharmacology 34, 2642–2654.

H.-H. Yeh et al. / NeuroImage 64 (2013) 630–639 Sankaranarayanapillai, M., Tong, W.P., Maxwell, D.S., Pal, A., Pang, J., Bornmann, W.G., Gelovani, J.G., Ronen, S.M., 2006. Detection of histone deacetylase inhibition by noninvasive magnetic resonance spectroscopy. Mol. Cancer Ther. 5, 1325–1334. Thomas, E.A., 2009. Focal nature of neurological disorders necessitates isotypeselective histone deacetylase (HDAC) inhibitors. Mol. Neurobiol. 40, 33–45. Tian, M., Ogawa, K., Wendt, R., Mukhopadhyay, U., Balatoni, J., Fukumitsu, N., Uthamanthil, R., Borne, A., Brammer, D., Jackson, J., Mawlawi, O., Yang, B., Alauddin, M.M., Gelovani, J.G., 2011. Whole-body biodistribution kinetics, metabolism, and radiation dosimetry estimates of 18F-PEG6-IPQA in nonhuman primates. J. Nucl. Med. 52, 934–941.

639

Tsankova, N.M., Berton, O., Renthal, W., Kumar, A., Neve, R.L., Nestler, E.J., 2006. Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action. Nat. Neurosci. 9, 519–525. Tsankova, N., Renthal, W., Kumar, A., Nestler, E.J., 2007. Epigenetic regulation in psychiatric disorders. Nat. Rev. Neurosci. 8, 355–367. Zeglis, B.M., Pillarsetty, N., Divilov, V., Blasberg, R.A., Lewis, J.S., 2011. The synthesis and evaluation of N1-(4-(2-[18F]-fluoroethyl)phenyl)-N8-hydroxyoctanediamide ([18F]FESAHA), a PET radiotracer designed for the delineation of histone deacetylase expression in cancer. Nucl. Med. Biol. 38, 683–696.