Chronic cocaine administration causes extensive white matter damagein brain: Diffusion tensor imaging and immunohistochemistry studies

Ponnada A. Narayana a,n, Juan J. Herrera a, Kurt H. Bockhorst a, Emilio Esparza-Coss a,Ying Xia a, Joel L. Steinberg b, F. Gerard Moeller b

a Department of Diagnostic and Interventional Imaging, University of Texas Health Science Center at Houston, Houston, TX 77030, USAb Department of Psychiatry, Virginia Commonwealth University School of Medicine, Richmond, VA, USA

a r t i c l e i n f o

Article history:Received 17 July 2013Received in revised form21 December 2013Accepted 14 January 2014Available online 23 January 2014

Keywords:Magnetic resonance imagingDiffusion tensor imagingCocaineRodentsImmunohistochemistryNeurobehavioral assay

a b s t r a c t

The effect of chronic cocaine exposure on multiple white matter structures in rodent brain was examinedusing diffusion tensor imaging (DTI), locomotor behavior, and end point histology. The animals receivedeither cocaine at a dose of 100 mg/kg (N¼19), or saline (N¼17) for 28 days through an implantedosmotic minipump. The animals underwent serial DTI scans, locomotor assessment, and end pointhistology for determining the expressions of myelin basic protein (MBP), neurofilament-heavy protein(NF-H), proteolipid protein (PLP), Nogo-A, aquaporin-4 (AQP-4), and growth associated protein-43 (GAP-43).Differences in the DTI measures were observed in the splenium (scc) and genu (gcc) of the corpus callosum(cc), fimbria (fi), and the internal capsule (ic). A significant increase in the activity in the fine motormovements and a significant decrease in the number of rearing events were observed in the cocaine-treated animals. Reduced MBP and Nogo-A and increased GAP-43 expressions were most consistentlyobserved in these structures. A decrease in the NF-H expression was observed in fi and ic. The reducedexpression of Nogo-A and the increased expression of GAP-43 may suggest destabilization of axonalconnectivity and increased neurite growth with aberrant connections. Increased GAP-43 suggests drug-induced plasticity or a possible repair mechanism response. The findings indicated that multiple whitematter tracts are affected following chronic cocaine exposure.

& 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Cocaine is a potent stimulant drug of abuse that can result inbehavioral and neurochemical changes in preclinical studies(Stankeviciute et al., 2013). In humans, cocaine abuse is alsoassociated with changes in brain structure and function (Moelleret al., 2005; Poon et al., 2007; Ma et al., 2009). Earlier neuro-chemical studies on the effects of cocaine abuse focused on thebrain gray matter (GM) structures. For example, it was shown thatmyelin proteins are significantly decreased in the nucleus accum-bens, which is implicated in addiction (Kovalevich et al., 2012).

Recent studies also found that cocaine affects white matter(WM) integrity. Diffusion tensor imaging (DTI) studies showedevidence of structural alterations in cocaine users in the corpuscallosum (cc) (Lim et al., 2002; Moeller et al., 2005, 2007; Ma et al.,2009; Lane et al., 2010). These impairments seen on DTI appear tobe associated with high levels of impulsivity, poor cognitive control,and mental flexibility (Moeller et al., 2005). Human studies, basedon DTI-derived transverse or radial diffusivity (RD) measurements,

have also suggested that chronic cocaine use may be associatedwith compromised myelin integrity (Moeller et al., 2007), consis-tent with mRNA and histology findings (Bannon et al., 2005).

Our controlled studies investigated if these changes in the DTImeasures observed in humans can be replicated in the rodent modelof cocaine exposure and determined the pathological underpinningsusing immunohistochemistry (Narayana et al., 2009). In our earlierstudy, based on the effects seen in human cocaine users, we focusedonly on the cc and demonstrated significant changes in the DTImeasures and altered expression in myelin basic protein (MBP) andneurofilament-heavy (NF-H) following chronic cocaine exposure.Previous studies also found similar changes in these proteins in thenucleus accumbens following cocaine administration (Kovalevichet al., 2012). Based on the mRNA studies, a robust and consistentdecrease in the expression of myelin-related genes, including MBP,proteolipid protein (PLP), and myelin-associated oligodendrocytebasic protein (MOBP), was observed in the nucleus accumbensfollowing cocaine exposure (Albertson et al., 2004). Some of thepublished studies also implicated other proteins such as Nogo-A andgrowth-associated protein-43 (GAP-43). Nogo-A is a protein that hasa central role in the inhibition of axonal growth (Chen et al., 2000;GrandPre et al., 2000). Nogo-A is highly expressed in oligodendro-cytes and in some neurons in the hippocampus, motor neurons, and

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/psychresns

Psychiatry Research: Neuroimaging

0925-4927/$ - see front matter & 2014 Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.pscychresns.2014.01.005

n Corresponding author. Tel.: þ1 713 500 7677.E-mail address: Ponnada.a.narayana@uth.tmc.edu (P.A. Narayana).

Psychiatry Research: Neuroimaging 221 (2014) 220–230

dorsal root ganglia (Huber et al., 2002; Schwab, 2002; Meier et al.,2003; Gil et al., 2006; Trifunovski et al., 2006). Nogo-A plays a role inrestricting plasticity in the central nervous system (McGee et al.,2005). Aquaporin-4 (AQP-4) is an integral membrane pore proteinthat mediates the movement of water (Preston and Agre, 1991) andhas been demonstrated to play a role in regulating extracellularcocaine-induced dopamine and glutamate release in brain (Li et al.,2006).

As indicated above, our previous study focused only on DTI andMBP and NF-H expressions in the cc in chronic cocaine adminis-tration (Narayana et al., 2009). However, based on some of thepublished studies referenced above, we hypothesized that chroniccocaine administration affects multiple WM structures, besidesthe cc, and a number of proteins. Therefore, the main purpose ofthe studies described in this article was to investigate the effect ofchronic cocaine exposure on (1) all the brain WM structures usingDTI; (2) the expressions of multiple WM proteins including MBP,PLP, NF-H, Nogo-A, GAP-43, and AQP-4 using immunohistochem-istry; and (3) motor behavior. A further goal of these studies was torelate the DTI changes to altered expressions in various proteins inrodents to gain an insight into cocaine-induced pathologicalchanges. In these studies we acquired DTI data with high spatialresolution (isotropic voxel of 0.27 mm) for reduced partial volumeaveraging.

2. Methods

2.1. Cocaine administration

Eight-week-old male Sprague-Dawley rats in the weight range of 280–300 g(289 g712.18 g) at baseline were included in this study. The animals were dividedinto the following two groups: (1) cocaine at a dose of 100 mg/kg (N¼17) and (2)saline-treated controls (N¼19). Cocaine dissolved in saline or saline alone (controlanimals) was continuously infused for 28 days through an Alzet osmotic minipump(Model 2ML4, Durect Corp., Cupertino, CA, USA), implanted subcutaneously asdescribed previously (Narayana et al., 2009). The infusion pump was filled with2 ml of 400 mg/ml of cocaine or 0.9% saline, which served as the control. The pumpdelivery rate was 2.5 ml/h. The pump was primed 24 h before implantation bywarming in a tube filled with saline at 37 1C in a water bath.

2.2. Magnetic resonance imaging (MRI) acquisition

All MRI scans were performed at baseline (before implantation of the infusionpump) and weekly thereafter for a period of 1 month. Thus, each animal wasscanned at five time points.

For MRI, animals were anesthetized with isoflurane at an induction dose of 4%and were placed on a custom-designed Plexiglas bed that was inserted into themagnet bore; the animals were secured with ear and tooth bars. During the scans,the animals were anesthetized with 2% isoflurane in a mixture of air (68%) andoxygen (30%) delivered by a rodent ventilator (Harvard Apparatus, South Natic, MA,USA) through a nose cone. The respiratory rate and rectal temperature weremonitored throughout the experiment with a physiologic monitoring unit (Model1025, SA Instruments, Stonybrook, NY, USA). A pulse oximeter (Model 570-100, SAInstruments, Stonybrook, NY, USA) was used to monitor the heart rate and oxygensaturation levels. An airflow warmer activated by a temperature probe (Model11007B, SA Instruments) maintained the body temperature at 36 1C.

A 7T/30 USR MRI scanner (Bruker BioSpin, Karlsruhe, Germany) with anactively shielded, water-cooled gradient coil system (Model BGA12; 116-mmdiameter) capable of producing maximum gradient amplitude of 400 mT m�1

with 80 μs rise time was used for image acquisition. The scanner operating systemwas ParaVision PV5.1. The vendor-supplied birdcage radiofrequency (RF) coil(Bruker BioSpin, Karlsruhe, Germany) with 72-mm internal diameter and 112-mm effective length was used for transmission. An in-house designed and remotelytunable receive-only RF coil (15-mm diameter) was placed over the head of theanimal and fixed to the Plexiglas bed. As a quality assurance procedure, before eachstudy, using a spin echo sequence, a homogenous water phantom doped with NiCl2was used to measure the signal-to-noise ratio (SNR) with the macro Auto_snr,which is part of the Bruker scanner software. In this method, the mean signal wasmeasured in a region of interest (ROI) of 3 by 3 voxels around the brightest voxels.The noise at the edges and corners of the image was measured. The lowest value(noise) in any of these locations was determined. The SNR was computed as meansignal/noise. In order to account for the effect of acquisition parameters such as

slice thickness, bandwidth, and number of averages, the Auto-snr routine calculatesthe SNR per unit volume as:

SNR=mm3 ¼ SNR � acqfactor� voxel factor

where voxel factor¼1/(volume in mm3) and acqfactor¼[256�5.12/(acqsize�acqtime)]�1, where acqtime is the sampling time in the frequency encodingdirection and acqsize is the number of phase encoding steps. The acqfactor is thenormalization factor that is calculated for standard acquisition of bandwidth of50 kHz and 256 complex points (20 ms/pt), and 256�256 (read�phase) matrixsize. This method yields SNR that is independent of the acquisition parameters andallows comparison across different scanners and with different acquisition para-meters. Even though in the current study all the acquisition parameters were keptthe same at different time points, we preferred using this method of computingSNR since it is calculated automatically and avoids any possible operator bias. Toconvert this to the commonly defined SNR in an image, the values reported hereshould be divided by 7.28.

Localized shimming over the brain was performed using a Bruker-suppliedField Map with adjustments up to third order shims followed by fine tuning of thefirst and second order shims using the PRESS sequence (echo time (TE)¼20 ms,repetition time (TR)¼2500 ms, TE1¼TE2¼10 ms, 2048 points with Hermite RFpulses) to consistently achieve water line width of less than 0.1 ppm (full width athalf maximum). A spin echo sequence with TE¼10 ms and TR¼5000 ms was usedto finely adjust the 901 RF pulse power manually over the central 1-mm axial slice(field of view (FOV)¼80 mm�80 mm, matrix¼128�128). The same geometrywas used for all other acquisitions with different sequences. For visualizing anyanatomical lesions, images were obtained using the dual echo 2D Rapid Acquisitionwith Relaxation Enhancement (RARE) sequence with the following parameters:number of slices¼28, slice thickness¼0.5 mm, FOV¼35 mm�35 mm, matrix¼256�256, effective TEs¼26 and 78 ms, TR¼5000 ms, Hermite pulse, RAREfactor¼4, spectral bandwidth¼60 KHz, number of dummy scans¼2, number ofaverages¼2. The total scan time was 10 min 40 s.

Diffusion-weighted images (DWIs) were acquired with a segmented (fourshots) 3D echoplanar imaging (EPI) sequence, using an icosahedral encodingscheme with bipolar gradients along 42 encoding directions (Madi et al., 2005).The other acquisition parameters were voxel size¼0.27 mm�0.27 mm�0.27 mm,FOV¼34.56 mm�20.00 mm�9.45 mm, matrix¼128�74�35, TE¼24.7 ms, TR¼500 ms, Hermite excitation RF pulse (pulse width¼3.6 ms, flip angle¼901),bandwidth¼250 kHz, number of averages¼1, and partial FT acceleration¼1.5.The EPI scan parameters included double sampling, automatic trajectory adjust-ment (that also turns on the navigator echoes), and regridding based on thetrajectory. The diffusion parameters were b-value¼800 s/mm2, diffusion gradientduration¼5 ms, diffusion gradient separation¼10 ms, and number of averageswithout any diffusion gradient (b0 images)¼9. The total scan time was 1 h.

2.3. MRI analysis

The RARE images were mainly used for visualizing anatomical lesions. Otherthan the routine processing that is a part of the scanner recon software, noadditional processing was performed on these images.

An in-house developed pipeline written in IDL software (Exelis Visual Informa-tion Solutions, Boulder, CO, USA) was used for automatically processing the DWIs.The pipeline included eddy current correction with the program ‘eddy_correct’,part of the FSL package (Smith et al., 2004), filtering noise and smoothing the data(Hahn et al., 2010), extracting brain using a semi-automatic method, and linear andnonlinear registration with automatic image registration (AIR; Woods et al., 1998a,1998b) to a template for group analysis. The unregistered images were exported toDtiStudio (Jiang et al., 2006), and the parametric maps of fractional anisotropy (FA),mean diffusivity (MD), RD, and longitudinal diffusivity (LD) were generated foreach animal.

Two levels of analysis were performed: (1) voxel based analysis of FA, and(2) ROI analysis. The voxel based analysis was performed using SPM8 (http://www.fil.ion.ucl.ac.uk/spm/software/) to identify voxels that differ in FA between controlsand cocaine animals. Since the only purpose of the SPM8 analysis was to guide inthe placement of the ROIs, we did not correct the SPM8 results for multiplecomparisons. The final results were based on the ROI analysis. The ROIs were drawnaround voxels that showed changes in the t-maps. Even with high isotropic spatialresolution of 0.27 mm, because of the small structure of the scc and gcc, we usedrelatively small ROIs to minimize partial volume averaging effects. The positioningof the ROIs was based on the 3D anatomy. The ROIs were drawn manually on eachslice for each animal by the same person, who was blind to the treatment. Giventhe experience of this person (about 15 years of experience in rat neuroanatomyand rodent brain MRI), it is unlikely that the operator bias in the ROI placement issignificant. However, such a bias cannot be completely ruled out. Since the sliceorientation varied slightly from animal to animal, the ROI size varied slightly (72pixels) to minimize partial volume averaging. For descriptive purposes, the meansand standard deviations of the parametric maps of the DTI measures within eachROI were calculated in each animal's native space using the ROI Manager of ImageJ(National Institute of Mental Health, Bethesda, MD, USA). Regions that did not show

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230 221

a high t-statistic were not included in the ROI analysis. The ROI analysis wasperformed on each animal (in native space).

2.4. Immunohistochemistry

Immunohistochemistry was performed only over the regions that showedsignificant differences (either false discovery rate corrected or uncorrected) in anyof the DTI measures (see below). On day 28 (end of cocaine delivery with the Alzetpump), following the last MRI scans, animals were transcardially perfused withsaline followed by 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS).The brains were removed and processed as described previously (Narayana et al.,2009). Brains were sectioned at 20 μm thick. Six brain sections, for staining witheach primary antibody, spanning the structure of interest from eight randomlyselected animals from each group, were chosen. These sections were labeled for thefollowing primary antibodies: MBP (1:1000; Covance, CA), NF-H (1:1000; Millipore,Billerica, MA), PLP (1:250; Abcam), Nogo-A (1:1000; Abcam), AQP-4 (1:1000;Chemicon), and GAP-43 (1:1000, Abcam). Primary antibodies were diluted withblocking solution (0.1 M PBS containing 5% goat serum and 0.3% Triton X-100).

Appropriate secondary antibodies were used at a dilution of 1:500 in 0.1 M PBS.The following Alexafluor dye-conjugated secondary antibodies were used: goatanti-mouse IgG Alexa Fluors 488 (Invitrogen, Carlsbad, CA, USA) and goat anti-rabbit IgG (HþL) Alex Fluors 568 (Invitrogen, Carlsbad, CA, USA). The regionsincluded for the immunohistochemistry were scc, gcc, and the body of splenium(bcc), fi, and ic (see Section 3).

2.5. Histology processing

All images were captured under 20� magnification using a Nikon Eclipse Ni-Emicroscope (Nikon Instruments, Melville, NY, USA). The exposure times weredetermined for each antibody in pilot experiments. The threshold levels were setto match those with positive staining and exclude background fluorescence.ImagePro Plus software (Media Cybernetics, Inc., Silver Spring, MD, USA) was usedto measure the percent areas or fluorescent intensity of the antibodies tested and inthe ROI as described previously (Herrera et al., 2008; Narayana et al., 2009). TheROIs were placed in scc, bcc, gcc, fi and ic. To ensure consistent analysis betweensections, the same ROI size (circular ROI with a 700-mm diameter) was maintainedfor each section throughout the experiment, and the threshold levels for eachprimary antibody were determined by random sampling from control sections.Personnel involved in the histologic analyses were blind to the treatment.

2.6. Neurobehavioral assays

Exploratory behavior that included fine motor movements, ambulation, andrearing activities was assessed using the computerized activity box to assess thegeneral motor behavior (Metz et al., 2000). In this assay animals were placed inthe activity chambers for a 30-min testing period, and the data were collected usingthe software and hardware provided by the Photobeam Activity System (PAS) withFlexfield (San Diego Instruments, Inc., San Diego, CA, USA). The PAS acquires dataon the animal's movements within the activity chamber by recording the numberof photo beams interrupted. The activity can be categorized as fine motor move-ments, gross movements, and rearing (McMahon and Cunningham, 2001). Baselinemonitoring was performed before the implantation of osmotic pumps, and weeklythereafter, approximately the same time of day for the duration of the study.

2.7. Statistical analysis

The ROI analysis of the proteins and of the DTI measures other than FA wereperformed using two-tailed Student's t-test between the control and cocaine-exposed groups. Statistical analysis was performed using GraphPad Prism (version5) software (GraphPad Software, San Diego, CA, USA). The significance level was setat 0.05 for all comparisons. A two-tailed Student's t-test was used to determinedifferences within each ROI of RD, LD, and the percent of expressions of MBP, NF-H,PLP, GAP-43, AQP-4, and Nogo-A between saline and cocaine-administered animals.Group behavioral data were analyzed using repeated measures one-way analysis ofvariance (ANOVA) followed by Bonferroni multiple comparison correction for thelocomotor analysis.

3. Results

3.1. Magnetic resonance imaging

The temporal stability of the SNR was excellent, except at a fewtime points (Fig. 1). No significant artifacts were observed on theimages. As can be seen in Fig. 1, low SNR was observed at two timepoints. We did not investigate the reasons for this low SNR. Datafrom animals on these 2 days were not excluded from the finalanalysis. We did not observe any lesions on the RARE images,indicating that cocaine administration did not result in any grosspathology. Typical T2-weighted RARE images and maps of DTI-derived measures are shown in Fig. 2.

3.1.1. DTI: voxel-based analysisAs an example, the results of the voxel-based analysis are

shown in Fig. 3. In this figure, voxels with differences in the FAvalues between controls and cocaine animals are shown in red(t-maps). These t-maps were generated with (left) and without(right) false discovery rate (FDR) correction for multiple compar-isons. With FDR correction, the only regions that showed signifi-cant differences (po0.05) were scc, gcc and ic. However,

Fig. 1. SNR as a function of time. Overall, the excellent temporal stability in the SNRcan be appreciated.

Fig. 2. Multi-modal MRI of 3 different brain regions (rows) from cocaine-exposed animals at the 4-week time point. Representative images of: RARE, fractional anisotropy (FA),mean diffusivity (MD), longitudinal diffusivity (LD), and radial diffusivity (RD).

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230222

differences in FA (po0.05) were seen in a few voxels in fi withoutthe FDR correction. It should be pointed out that the t-maps showmore extensive injury in the left hemisphere of the brain com-pared with the right. However, the reasons for this observedasymmetry are unclear.

3.1.2. ROI analysisThe ROI placement was guided by the t-maps, which were not

corrected for multiple comparisons (Fig. 4 left). The scc and gcc arevery small structures in rats. Even though our spatial resolution of0.27 mm isotropic voxel is one of the highest reported for in vivo

Fig. 4. Representative images showing the placement of the ROIs for comparison between cocaine-treated and saline control animals. The ROIs were placed in the (A) genu,(B) fimbria, (C) internal capsule, and (D) splenium.

Fig. 3. The t-maps (shown in red) with significantly different FA values between controls and cocaine-treated animals superimposed on the RARE images, without (left) andwith (right) FDR correction for multiple comparison. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of thisarticle.)

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230 223

Fig. 5. Results of the ROI analysis of the DTI measures for the scc, gcc, bcc, ic, and fi. The p values are also shown in this figure. The error bars represent the standarddeviation.

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230224

rodent studies, we still had to use a small number of voxels in agiven slice to balance between a small number of voxels, whichcould result in statistical noise, and a large number of voxels,which would introduce partial volume averaging. However, theROI extended to other slices. For ic and fi, ROIs were placed in bothhemispheres. The ROI for scc consisted of 28 voxels (0.55 mm3),gcc 71 voxels (1.4 mm3), fi 72 voxels (1.42 mm3), and ic 75 voxels(1.48 mm3). As pointed out earlier, the protein analysis anddescriptive DTI measures within each ROI were performed in thenative space of each animal and thus did not involve registrationerrors.

Increased RD was observed in scc and fi (p¼0.0321). A strongtrend toward an increase was observed for RD in gcc (p¼0.0576).LD decreased in scc (p¼0.0339), gcc (p¼0.0002), and ic(p¼0.0084). The results of the DTI measures within each ROI aresummarized in Fig. 5 and Table 1. The error bars in Fig. 5 representthe standard deviation.

3.2. Immunohistochemistry

We matched the histology and DTI sections for the ROI analysisas closely as possible by visual inspection, realizing that these twomodalities greatly differ in spatial resolution. As an example, Fig. 6shows the matched histology and DTI sections. As an example,representative immunofluorescence images of fimbria from thecontrol and cocaine-treated brains are shown in Fig. 7.

3.2.1. Corpus callosum3.2.1.1. scc. Immunofluorescent quantification indicated significantlylower MBP (p¼0.0007) and Nogo-A (p¼0.0002) in the cocaine-treated animals than in the control animals. The cocaine-treatedanimals showed significantly higher GAP-43 expression (po0.0001)compared with the saline controls. A trend toward a significantlyhigher expression of NF-H isoform (p¼0.0506) and PLP (p¼0.0801)was seen in cocaine-treated animals. No significant difference in theexpression of AQP-4 (p¼0.2861) was observed between the twogroups. The results of quantitative analysis of the immuno-fluorescence for scc are shown in Fig. 8 (where the error barsrepresent the standard deviation).

3.2.1.2. gcc. The cocaine-treated rats showed significantly lowerNogo-A (po0.0001), and higher expression of GAP-43 (p¼0.0151).A trend toward a decrease was observed in the expression of MBP(p¼0.0637). Cocaine-treated rats showed higher NF-H expression,

but this difference was not statistically significant (p¼0.2299). Nosignificant differences were observed in the expressions of PLP (p¼0.3936) and AQP-4 (p¼0.3581).

3.2.1.3. bcc. We observed no significant differences in MBP (p¼0.3282) and Nogo-A (p¼0.8913). Similarly, we did not observesignificant differences in NF-H (p¼0.4314) and AQP-4 (p¼0.9162)expression. Cocaine-treated rats did, however, show significantlyhigher expression of GAP-43 (p¼0.0445) and a trend towardstatistical significance for PLP (p¼0.0712).

Table 1Effect of cocaine treatment on the DTI measures compared with saline. The numerical values represent mean7S.D. and the corresponding p values are shown in parenthesis.

DTI measures Treatment White matter structures

scc bcc gcc fi ic

FA (dimensionless) Saline 0.72770.018 0.61270.103 0.75270.028 0.71670.045 0.66170.042Cocaine 0.68970.040 0.58970.077 0.71970.028 0.73170.039 0.62770.036

(0.0027)# (0.4994) (0.0039)# (0.2889) (0.0119)#

MD (mm2 s�1�10�6) Saline 7987250 7897268 8907279 9107204 671796Cocaine 715790 6827146 7467140 8177182 6977149

(0.2231) (0.1786) (0.0779) (0.1580) (0.5422)

LD (mm2 s�1�10�6) Saline 1797765 1435792 1870757 19217111 1537783Cocaine 1732791 1416799 1775764 1959790 1482760

(0.0339)# (0.5973) (0.0002)# (0.2651) (0.0280)#

RD mm2 s�1�10�6) Saline 415730 4837268 397740 460757 441746Cocaine 449742 513799 421739 446754 466737

(0.0124)n (0.6746) (0.0576) (0.4301) (0.0806)

# and n denote significant decrease and increase (at pr0.05), respectively.

Fig. 6. Representative matched histology and DTI sections in cocaine-treatedanimals. The matching can be seen even with dissimilar spatial resolution.

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230 225

3.2.2. fi and icWe observed a significant decrease in the expression of MBP in

both fi (po0.0001) and ic (p¼0.001) in cocaine-treated animalscompared with controls. Also in the cocaine-treated animals there

was a significant decrease in the expression of NF-H in fi

(p¼0.0007) and ic (p¼0.0015). An increase in axonal plasticity,as measured by GAP-43, was observed in both fi (po0.0001) andic (p¼0.0011). There was a significant decrease in the expression

Fig. 7. Representative images of brain sections of the fimbria from control and cocaine-exposed brains. The images were obtained to demonstrate the labeling patterns of[panels A and B] myelin basic protein (MBP), [C and D] Nogo-A, [E and F] GAP-43, [G and H] neurofilament – heavy isoform (NF-H), [J and K] proteolipid protein (PLP), and[L and M] aquaporin-4 (AQP-4). scc¼splenium. Scale bars¼100 mm.

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230226

of Nogo-A in both fi (po0.0001) and ic (po0.0001). No significantdifference was observed in the expression of PLP in either the fi

(p¼0.5102) or the ic (p¼0.7228). AQP-4 expression was notdetermined in these regions.

The results of the quantification of immunohistochemicalanalysis along with the p values for different structures aresummarized in Table 2.

3.3. Neurobehavioral assessment

Behavioral assessment indicated a significant increase in thecocaine-treated animals in the activity counts in the fine motormovements compared to baseline measures (Fig. 9A). A significantdecrease was observed in the number of rearing events in thecocaine-treated animals (Fig. 9B). Both these changes were seen at

Fig. 8. Quantitative analysis of the expressions of (A) MBP, (B) Nogo-A, (C) GAP-43, (D) NF-H, (E) PLP, and (F) AQP-4 in the splenium of the corpus callosum. A significantdecrease was observed in the markers for oligodendrocytes (A and B). A significant increase was observed in GAP-43, a marker of neuronal plasticity. A trend towardsstatistically significant increase in NF-H was also determined (D). No significant difference was observed in PLP, and AQP-4 (F). Error bars represent the standard deviation.

Table 2Effect of cocaine treatment on protein measures compared with saline. The numerical values represent mean7S.D. and the corresponding p values are shown inparentheses.

Protein Treatment White matter structures

scc bcc gcc ic fi

MBP Saline 0.019270.02 29.96721.99 2.04471.80 7.91677.77 16.78713.47Cocaine 0.005370.01 34.13719.53 1.43171.34 5.92476.29 6.31375.39

(0.0007)# (0.3282) (0.0637) (0.001)# (o0.0001)#

Nogo-A Saline 0.785271.04 5.924711.86 0.769270.60 0.948271.19 21.20716.50Cocaine 0.164170.31 6.260712.15 0.249470.47 0.459270.59 9.185712.20

(0.0002)# (0.8913) (o0.0001)# (o0.0001)# (o0.0001)#

GAP-43 Saline 1.72272.54 20.25728.20 2.23372.026 1.28671.91 14.6378.82Cocaine 6.20275.45 32.94732.65 3.85473.90 1.88272.23 25.54712.61

(o0.0001)n (0.0445)n (0.0151)n (o0.0001)n (0.0011)n

NF-H Saline 7.9176.35 32.29724.14 1.10070.97 4.21775.81 7.05876.72Cocaine 10.6476.57 35.96721.27 1.35771.05 2.73974.74 4.28273.95

(0.0506) (0.4314) (0.2299) (0.0007)# (0.0015)#

PLP Saline 1.55471.38 17.80718.08 1.23971.06 3.48775.06 15.56714.43Cocaine 1.09171.08 24.29716.74 1.06770.85 3.34673.70 14.23711.22

(0.0801) (0.0712) (0.3963) (0.7228) (0.5102)

AQP-4 Saline 2.96772.04 5.94274.61 0.98470.70 ND NDCocaine 3.39471.74 6.05375.71 0.857170.63

(0.2861) (0.9162) (0.3581)

# and n denote significant decrease and increase (at pr0.05), respectively.ND¼not determined.

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230 227

3 weeks after the start of cocaine treatment. No significantdifferences in the ambulatory exploration between the controland cocaine-administered animals were observed.

4. Discussion

We believe that this is the first comprehensive study toinvestigate the effects of cocaine using DTI, neurobehavioral assaybased on computer activity boxes, and multiple immunohisto-chemical stains. We observed altered protein expressions in multi-ple WM structures after chronic administration of cocaine, anddifferences in DTI metrics, especially RD, in these same brainregions (summarized in Table 1). In addition to confirming ourearlier results that scc was affected in cocaine-treated animals(Narayana et al., 2009), other WM structures such as fi and ic werealso affected by chronic cocaine administration. The implicationsof these observations are discussed below.

The DTI data indicated differences between cocaine- andsaline-treated rodents in the cc and the ic. In both of thesestructures, we observed decreased FA, consistent with microstruc-tural changes in the white matter. The reduced MBP expressionwould suggest compromised myelin, which is consistent withincreased RD in the scc and ic (Song et al., 2005; Herrera et al.,2008; Budde et al., 2009; Narayana et al., 2009). In the case of thefi, reduced MBP expression was observed even though a change inRD was not evident. Still, in majority of the structures, concomi-tant changes in both DTI and histological measures were observed.

The LD results were less consistently associated with histo-pathology across different structures. For example, a decrease inLD was observed in the genu and the splenium of the cc and ic.However, a trend toward an increase in NF-H was observed in thescc (p¼0.0506). This lack of association between LD and NF-Hexpression is consistent with more recent reports (Budde et al.,2007, 2011; DeBoy et al., 2007; Herrera et al., 2008), whichindicate that the original concept of LD being solely related toaxonal injury might be overly simplistic.

A robust observation on DTI was increased RD in the scc, alongwith a significant reduction in MBP in the same brain region. TheDTI findings are consistent with our previous rodent study and

multiple previous human studies in cocaine use (Moeller et al.,2005, 2007; Narayana et al., 2009). Thus, RD could serve as abiomarker of WM structural damage in cocaine addiction.

The present study demonstrated significantly lower MBP andNogo-A expression in the scc, gcc, fi and ic, suggesting that myelincompromise extends to multiple WM structures. Among the threesubstructures of the cc, the splenium was most affected, followedby the genu. The reasons for this differential effect across thesubstructures of the cc are not completely clear. It was reportedthat the percentage of myelinated axons showing ultrastructuralpathological changes with aging in male Sprague-Dawley rats ismuch higher in the scc compared with the gcc and the bcc (Sargonet al., 2007). It was also shown that in this species the number ofmyelinated axons is higher in the scc compared with the gcc andthe bcc (Sargon et al., 2003). Recent studies also suggest that theaxonal caliber is much higher in the bcc compared with the gccand the scc (Barazany et al., 2009). Perhaps, this suggests thatlarger axons are less vulnerable to cocaine exposure. In addition,there may be differential expression of proteins involved in lipidperoxidation, oxidative stress, and apoptosis across the cc sub-regions (Kashem et al., 2009).

The ic is also affected by cocaine exposure. Myelin-relatedtranscripts encoding MBP, cyclic nucleotide 3-phosphodiesterase,MAG, and MOG are decreased following cocaine exposure(Kristiansen et al., 2009). The change in the ic was also reportedon MRI in cocaine subjects (Lim et al., 2008). White matter fibersin the ic traverse through the basal ganglia allowing connectionsbetween the cortex, brainstem, spinal cord, and thalamus. Damageto the ic has been demonstrated to correlate with motor impair-ment in stroke (Pendlebury et al., 1999) and multiple sclerosis (Leeet al., 2000). Our neurobehavorial data also suggest impairment inrearing activity, which may indicate damage to the corticospinaltract that traverses through posterior limb of the ic.

The fimbria is a major route for afferent and efferent fibers ofthe hippocampal formation. It plays a critical role as a pathwaybetween the hippocampus and nucleus accumbens, possibly inregulating cocaine-seeking behavior (Sabeti et al., 2003; Russoet al., 2010; Kovalevich et al., 2012). Published studies indicate thatglutamate excitotoxicity induces apoptosis in oligodendrocytesthrough an increased caspase-3 cleavage (McDonald et al., 1998;Sanchez-Gomez et al., 2011; Kovalevich et al., 2012). One mightspeculate that the altered state of myelin observed in the fi reflectschanges occurring in the glutamatergic pathway.

An interesting observation in our study is the reduced expres-sion of Nogo-A. Reduced Nogo-A promotes neuronal sproutingthat could be beneficial in traumatic injury (Liebscher et al., 2005)and early stages of Alzheimer's disease, by compensating for thesynaptic loss. However, in the later part of the disease, thesprouting axons may cause aberrant connections that result infurther neurodegeneration (Hashimoto and Masliah, 2003). Nogo-A is also shown to stabilize axonal connectivity (Park et al., 2010;Schwab, 2010).

Our findings of reduced Nogo-A are similar to the results ofanother study that examined WM changes following ampheta-mine exposure (Yang et al., 2011). That study demonstrated asignificant decrease in the oligodendrocyte marker MBP, Nogo-A,and glutathione S-transferase in the frontal cortex and cc inanimals treated with amphetamine, suggesting that differentstimulants may exert similar effects on myelin-related proteins.

Overall, chronic cocaine exposure produces several changes invarious proteins that ultimately lead to WM injury. One potentialmechanism through which cocaine could exert its damaging effectson WM is oxidative stress. Cocaine exposure induces oxidative stresswithout apoptosis in the rodent frontal cortex and striatum (Dietrichet al., 2005), increases the amount of dopamine in the synaptic cleft,which is metabolized by auto-oxidation or monoamine oxidase, thus

Fig. 9. Motor responses in animals treated with cocaine were compared withbaseline measures using computerized activity box monitoring. (A) A significantincrease in fine motor movements was observed 3 weeks after chronic cocaineadministration compared with baseline measures (p¼0.0017). (B) A significantreduction in the number of rearing events was also observed at 3 and 4 weeks ofchronic cocaine administration compared with baseline measures (p¼0.0049 and0.0027, respectively).

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230228

increasing reactive oxygen species (ROS) production (Hermida-Ameijeiras et al., 2004). A recent study indicated that Nogo-A hasneuroprotective effects against oxidative stress (Mi et al., 2012).Amino-Nogo-A was demonstrated to scavenge ROS by interactingwith oxidized peroxiredoxin 2 in vitro (Mi et al., 2012). Oxidativestress has been demonstrated to precede death of human progenitorcells following cocaine exposure in vitro (Poon et al., 2007). Knock-down of Nogo-A may also cause acute oxidative stress and markedlyincrease neuronal death (Mi et al., 2012). The possible increasedproduction of ROS combined with the decreased Nogo-A suggeststhat oxidative stress may be a contributing factor in the pathologyobserved in our study.

The present study demonstrated a significant increase in theexpression of GAP-43. GAP-43 plays a key role in guiding thegrowth of axons and modulating new connections and enablesneurons to sprout new terminals (Aigner et al., 1995). Recentstudies demonstrated that a decrease in Nogo-A resulted in anincrease in axonal plasticity as determined by GAP-43 expression(Dietrich et al., 2005; Masliah et al., 2010). Increased expression ofGAP-43 was also observed in the fimbria in animals that wereprenatally administered cocaine (Clarke et al., 1996). In our studies,changes in Nogo-A expressionwere observed in some structures inthe absence of changes in MBP and NF-H. This perhaps suggeststhat GAP-43 is an early indicator of cocaine-induced pathologicalchanges.

It was suggested that late developing WM may be more vulner-able to increased dopamine following amphetamine exposure (Yanget al., 2011). Cocaine administration also produces an increase in theextracellular dopamine levels, and changes in WM pathology havebeen observed in areas away from the cc (Hermida-Ameijeiras et al.,2004; Masliah et al., 2010). Oligodendrocytes express D2 and D3receptors for dopamine, and stimulation of these receptors resultedin decreased transition of immature oligodendrocytes to matureoligodendrocytes (Bongarzone et al., 1998). In addition to dopaminerelease, cocaine administration results in alteration of glutamateneurotransmission leading to increased glutamate levels in variousbrain regions that could alter protein expressions in cc. A follow-upstudy is needed to examine the glutamate levels and the expressionof glutamate receptors in cc.

It was demonstrated that repeated chronic administrations ofcocaine resulted in locomotor sensitization (Gulley et al., 2003; Sabetiet al., 2003). The present study has shown that chronic administra-tion of cocaine resulted in an increase in fine motor movements anda decrease in rearing events compared with baseline measures.Increases in fine motor movements were also observed in anotherstudy (Wiley et al., 2008). We speculate that the increase in the finemotor movements may be related to the increase in the plasticity inthe cc, but this possibility needs to be examined in a follow-up study.Decreases in rearing events were observed in another chroniccocaine study (van Haaren and Meyer, 1991). The present study alsoshowed no significant difference between groups in the expression ofAQP-4, which has been suggested to be affected in cocaine addiction(Li et al., 2006).

In conclusion, this study extends previous investigation show-ing an effect of chronic cocaine on brain WM matter structure. Thefindings of reduced expression of Nogo-A and increased levels ofGAP-43 could destabilize axonal connectivity and increased neuritegrowth with aberrant connections. Increased GAP-43 perhapssuggests stimulant-induced plasticity or a repair mechanismresponse. Follow-up studies are needed to gain a better under-standing of the cellular and molecular mechanisms underlying thecocaine-induced loss of WM protein. Understanding cocaine-induced neuronal plasticity may be critical for identifying noveldrug targets for blocking or reversing cocaine-induced changes inbrain structure and associated behavioral problems.

Acknowledgments

This work is supported National Institute on Drug Abuse (NIDA)Grant #P50 DA009262. NIDA is not responsible for the contents ofthis article. The 7T scanner used in this study was funded by NIHGrant #S10 RR17205. We thank Mark Mattingly from Bruker forvery useful discussions about the SNR measurements.

References

Aigner, L., Arber, S., Kapfhammer, J.P., Laux, T., Schneider, C., Botteri, F., Brenner, H.R.,Caroni, P., 1995. Overexpression of the neural growth-associated protein GAP-43induces nerve sprouting in the adult nervous system of transgenic mice. Cell 83,269–278.

Albertson, D.N., Pruetz, B., Schmidt, C.J., Kuhn, D.M., Kapatos, G., Bannon, M.J., 2004.Gene expression profile of the nucleus accumbens of human cocaine abusers:evidence for dysregulation of myelin. Journal of Neurochemistry 88, 1211–1219.

Bannon, M., Kapatos, G., Albertson, D., 2005. Gene expression profiling in the brainsof human cocaine abusers. Addiction Biology 10, 119–126.

Barazany, D., Basser, P.J., Assaf, Y., 2009. In vivo measurement of axon diameterdistribution in the corpus callosum of rat brain. Brain 132, 1210–1220.

Bongarzone, E.R., Howard, S.G., Schonmann, V., Campagnoni, A.T., 1998. Identifica-tion of the dopamine D3 receptor in oligodendrocyte precursors: potential rolein regulating differentiation and myelin formation. Journal of Neuroscience 18,5344–5353.

Budde, M.D., Janes, L., Gold, E., Turtzo, L.C., Frank, J.A., 2011. The contribution ofgliosis to diffusion tensor anisotropy and tractography following traumaticbrain injury: validation in the rat using Fourier analysis of stained tissuesections. Brain 134, 2248–2260.

Budde, M.D., Kim, J.H., Liang, H.F., Schmidt, R.E., Russell, J.H., Cross, A.H., Song, S.K.,2007. Toward accurate diagnosis of white matter pathology using diffusiontensor imaging. Magnetic Resonance in Medicine 57, 688–695.

Budde, M.D., Xie, M., Cross, A.H., Song, S.K., 2009. Axial diffusivity is the primarycorrelate of axonal injury in the experimental autoimmune encephalomyelitisspinal cord: a quantitative pixelwise analysis. Journal of Neuroscience 29,2805–2813.

Chen, M.S., Huber, A.B., van der Haar, M.E., Frank, M., Schnell, L., Spillmann, A.A.,Christ, F., Schwab, M.E., 2000. Nogo-A is a myelin-associated neurite outgrowthinhibitor and an antigen for monoclonal antibody IN-1. Nature 403, 434–439.

Clarke, C., Clarke, K., Muneyyirci, J., Azmitia, E., Whitaker-Azmitia, P.M., 1996.Prenatal cocaine delays astroglial maturation: immunodensitometry showsincreased markers of immaturity (vimentin and GAP-43) and decreasedproliferation and production of the growth factor S-100. Developmental BrainResearch 91, 268–273.

DeBoy, C.A., Zhang, J., Dike, S., Shats, I., Jones, M., Reich, D.S., Mori, S., Nguyen, T.,Rothstein, B., Miller, R.H., Griffin, J.T., Kerr, D.A., Calabresi, P.A., 2007. Highresolution diffusion tensor imaging of axonal damage in focal inflammatoryand demyelinating lesions in rat spinal cord. Brain 130, 2199–2210.

Dietrich, J.B., Mangeol, A., Revel, M.O., Burgun, C., Aunis, D., Zwiller, J., 2005. Acuteor repeated cocaine administration generates reactive oxygen species andinduces antioxidant enzyme activity in dopaminergic rat brain structures.Neuropharmacology 48, 965–974.

Gil, V., Nicolas, O., Mingorance, A., Urena, J.M., Tang, B.L., Hirata, T., Saez-Valero, J.,Ferrer, I., Soriano, E., del Rio, J.A., 2006. Nogo-A expression in the humanhippocampus in normal aging and in Alzheimer disease. Journal of Neuro-pathology and Experimental Neurology 65, 433–444.

GrandPre, T., Nakamura, F., Vartanian, T., Strittmatter, S.M., 2000. Identification ofthe Nogo inhibitor of axon regeneration as a Reticulon protein. Nature 403,439–444.

Gulley, J.M., Hoover, B.R., Larson, G.A., Zahniser, N.R., 2003. Individual differences incocaine-induced locomotor activity in rats: behavioral characteristics, cocainepharmacokinetics, and the dopamine transporter. Neuropsychopharmacology28, 2089–2101.

Hahn, K., Prigarin, S., Rodenacker, K., Hasan, K., 2010. Denoising for diffusion tensorimaging with low signal to noise ratios: method and Monte Carlo validation.International Journal for Biomathematics and Biostatistics 1, 63–81.

Hashimoto, M., Masliah, E., 2003. Cycles of aberrant synaptic sprouting andneurodegeneration in Alzheimer’s and dementia with Lewy bodies. Neuro-chemical Research 28, 1743–1756.

Hermida-Ameijeiras, A., Mendez-Alvarez, E., Sanchez-Iglesias, S., Sanmartin-Suarez, C.,Soto-Otero, R., 2004. Autoxidation and MAO-mediated metabolism of dopamineas a potential cause of oxidative stress: role of ferrous and ferric ions. Neuro-chemistry International 45, 103–116.

Herrera, J.J., Chacko, T., Narayana, P.A., 2008. Histological correlation of diffusiontensor imaging metrics in experimental spinal cord injury. Journal of Neuro-science Research 86, 443–447.

Huber, A.B., Weinmann, O., Brosamle, C., Oertle, T., Schwab, M.E., 2002. Patterns ofNogo mRNA and protein expression in the developing and adult rat and afterCNS lesions. Journal of Neuroscience 22, 3553–3567.

Jiang, H., van Zijl, P.C., Kim, J., Pearlson, G.D., Mori, S., 2006. DtiStudio: resourceprogram for diffusion tensor computation and fiber bundle tracking. ComputerMethods and Programs in Biomedicine 81, 106–116.

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230 229

Kashem, M.A., Etages, H.D., Kopitar-Jerala, N., McGregor, I.S., Matsumoto, I., 2009.Differential protein expression in the corpus callosum (body) of humanalcoholic brain. Journal of Neurochemistry 110, 486–495.

Kovalevich, J., Corley, G., Yen, W., Rawls, S.M., Langford, D., 2012. Cocaine-inducedloss of white matter proteins in the adult mouse nucleus accumbens isattenuated by administration of a beta-lactam antibiotic during cocaine with-drawal. American Journal of Pathology 181, 1921–1927.

Kristiansen, L.V., Bannon, M.J., Meador-Woodruff, J.H., 2009. Expression of tran-scripts for myelin related genes in postmortem brain from cocaine abusers.Neurochemical Research 34, 46–54.

Lane, S.D., Steinberg, J.L., Ma, L., Hasan, K.M., Kramer, L.A., Zuniga, E.A., Narayana, P.A.,Moeller, F.G., 2010. Diffusion tensor imaging and decision making in cocainedependence. PLoS One 5, e11591.

Lee, M.A., Blamire, A.M., Pendlebury, S., Ho, K.H., Mills, K.R., Styles, P., Palace, J.,Matthews, P.M., 2000. Axonal injury or loss in the internal capsule and motorimpairment in multiple sclerosis. Archives of Neurology 57, 65–70.

Li, Z., Gao, L., Liu, Q., Cao, C., Sun, X.L., Ding, J.H., Hu, G., 2006. Aquaporin-4 knockoutregulated cocaine-induced behavior and neurochemical changes in mice.Neuroscience Letters 403, 294–298.

Liebscher, T., Schnell, L., Schnell, D., Scholl, J., Schneider, R., Gullo, M., Fouad, K., Mir,A., Rausch, M., Kindler, D., Hamers, F.P., Schwab, M.E., 2005. Nogo-A antibodyimproves regeneration and locomotion of spinal cord-injured rats. Annals ofNeurology 58, 706–719.

Lim, K.O., Choi, S.J., Pomara, N., Wolkin, A., Rotrosen, J.P., 2002. Reduced frontalwhite matter integrity in cocaine dependence: a controlled diffusion tensorimaging study. Biological Psychiatry 51, 890–895.

Lim, K.O., Wozniak, J.R., Mueller, B.A., Franc, D.T., Specker, S.M., Rodriguez, C.P.,Silverman, A.B., Rotrosen, J.P., 2008. Brain macrostructural and microstructuralabnormalities in cocaine dependence. Drug and Alcohol Dependence 92,164–172.

Ma, L., Hasan, K.M., Steinberg, J.L., Narayana, P.A., Lane, S.D., Zuniga, E.A., Kramer, L.A.,Moeller, F.G., 2009. Diffusion tensor imaging in cocaine dependence: regionaleffects of cocaine on corpus callosum and effect of cocaine administration route.Drug and Alcohol Dependence 104, 262–267.

Madi, S., Hasan, K.M., Narayana, P.A., 2005. Diffusion tensor imaging of in vivo andexcised rat spinal cord at 7T with an icosahedral encoding scheme. MagneticResonance in Medicine 53, 118–125.

Masliah, E., Xie, F., Dayan, S., Rockenstein, E., Mante, M., Adame, A., Patrick, C.M.,Chan, A.F., Zheng, B., 2010. Genetic deletion of Nogo/Rtn4 ameliorates beha-vioral and neuropathological outcomes in amyloid precursor protein transgenicmice. Neuroscience 169, 488–494.

McDonald, J.W., Althomsons, S.P., Hyrc, K.L., Choi, D.W., Goldberg, M.P., 1998.Oligodendrocytes from forebrain are highly vulnerable to AMPA/kainatereceptor-mediated excitotoxicity. Nature Medicine 4, 291–297.

McGee, A.W., Yang, Y., Fischer, Q.S., Daw, N.W., Strittmatter, S.M., 2005. Experience-driven plasticity of visual cortex limited by myelin and Nogo receptor. Science309, 2222–2226.

McMahon, L.R., Cunningham, K.A., 2001. Antagonism of 5-hydroxytryptamine(2a)receptors attenuates the behavioral effects of cocaine in rats. Journal ofPharmacology and Experimental Therapeutics 297, 357–363.

Meier, S., Brauer, A.U., Heimrich, B., Schwab, M.E., Nitsch, R., Savaskan, N.E., 2003.Molecular analysis of Nogo expression in the hippocampus during develop-ment and following lesion and seizure. FASEB Journal: Official Publication ofthe Federation of American Societies for Experimental Biology 17, 1153–1155.

Metz, G.A., Merkler, D., Dietz, V., Schwab, M.E., Fouad, K., 2000. Efficient testing ofmotor function in spinal cord injured rats. Brain Research 883, 165–177.

Mi, Y.J., Hou, B., Liao, Q.M., Ma, Y., Luo, Q., Dai, Y.K., Ju, G., Jin, W.L., 2012. Amino-Nogo-A antagonizes reactive oxygen species generation and protects immatureprimary cortical neurons from oxidative toxicity. Cell Death and Differentiation19, 1175–1186.

Moeller, F.G., Hasan, K.M., Steinberg, J.L., Kramer, L.A., Dougherty, D.M., Santos, R.M.,Valdes, I., Swann, A.C., Barratt, E.S., Narayana, P.A., 2005. Reduced anteriorcorpus callosum white matter integrity is related to increased impulsivity andreduced discriminability in cocaine-dependent subjects: diffusion tensor ima-ging. Neuropsychopharmacology 30, 610–617.

Moeller, F.G., Hasan, K.M., Steinberg, J.L., Kramer, L.A., Valdes, I., Lai, L.Y., Swann, A.C.,Narayana, P.A., 2007. Diffusion tensor imaging eigenvalues: preliminary evidence

for altered myelin in cocaine dependence. Psychiatry Research: Neuroimaging154, 253–258.

Narayana, P.A., Ahobila-Vajjula, P., Ramu, J., Herrera, J., Steinberg, J.L., Moeller, F.G.,2009. Diffusion tensor imaging of cocaine-treated rodents. Psychiatry Research:Neuroimaging 171, 242–251.

Park, K.J., Grosso, C.A., Aubert, I., Kaplan, D.R., Miller, F.D., 2010. p75NTR-dependent,myelin-mediated axonal degeneration regulates neural connectivity in theadult brain. Nature Neuroscience 13, 559–566.

Pendlebury, S.T., Blamire, A.M., Lee, M.A., Styles, P., Matthews, P.M., 1999. Axonalinjury in the internal capsule correlates with motor impairment after stroke.Stroke 30, 956–962.

Poon, H.F., Abdullah, L., Mullan, M.A., Mullan, M.J., Crawford, F.C., 2007. Cocaine-induced oxidative stress precedes cell death in human neuronal progenitorcells. Neurochemistry International 50, 69–73.

Preston, G.M., Agre, P., 1991. Isolation of the cDNA for erythrocyte integralmembrane protein of 28 kilodaltons: member of an ancient channel family.Proceedings of the National Academy of Sciences of the United States ofAmerica 88, 11110–11114.

Russo, S.J., Dietz, D.M., Dumitriu, D., Morrison, J.H., Malenka, R.C., Nestler, E.J., 2010.The addicted synapse: mechanisms of synaptic and structural plasticity innucleus accumbens. Trends in Neurosciences 33, 267–276.

Sabeti, J., Gerhardt, G.A., Zahniser, N.R., 2003. Individual differences in cocaine-induced locomotor sensitization in low and high cocaine locomotor-respondingrats are associated with differential inhibition of dopamine clearance in nucleusaccumbens. Journal of Pharmacology and Experimental Therapeutics 305,180–190.

Sanchez-Gomez, M.V., Alberdi, E., Perez-Navarro, E., Alberch, J., Matute, C., 2011. Baxand calpain mediate excitotoxic oligodendrocyte death induced by activation ofboth AMPA and kainate receptors. Journal of Neuroscience 31, 2996–3006.

Sargon, M.F., Denk, C.C., Celik, H.H., Surucu, H.S., Aldur, M.M., 2007. Electronmicroscopic examination of the myelinated axons of corpus callosum inperfused young and old rats. International Journal of Neuroscience 117,999–1010.

Sargon, M.F., Mas, N., Senan, S., Ozdemir, B., Celik, H.H., Cumhur, M., 2003.Quantitative analysis of myelinated axons of commissural fibers in the ratbrain. Anatomia, Histologia, Embryologia 32, 141–144.

Schwab, M.E., 2002. Increasing plasticity and functional recovery of the lesionedspinal cord. Progress in Brain Research 137, 351–359.

Schwab, M.E., 2010. Functions of Nogo proteins and their receptors in the nervoussystem. Nature Reviews of Neuroscience 11, 799–811.

Smith, S.M., Jenkinson, M., Woolrich, M.W., Beckmann, C.F., Behrens, T.E., Johansen-Berg, H., Bannister, P.R., De Luca, M., Drobnjak, I., Flitney, D.E., Niazy, R.K.,Saunders, J., Vickers, J., Zhang, Y., De Stefano, N., Brady, J.M., Matthews, P.M.,2004. Advances in functional and structural MR image analysis and imple-mentation as FSL. Neuroimage 23 (Suppl 1), S208–S219.

Song, S.K., Yoshino, J., Le, T.Q., Lin, S.J., Sun, S.W., Cross, A.H., Armstrong, R.C., 2005.Demyelination increases radial diffusivity in corpus callosum of mouse brain.Neuroimage 26, 132–140.

Stankeviciute, N.M., Scofield, M.D., Kalivas, P.W., Gipson, C.D., 2013. Rapid, transientpotentiation of dendritic spines in context-induced relapse to cocaine seeking.Addiction Biology.

Trifunovski, A., Josephson, A., Bickford, P.C., Olson, L., Brene, S., 2006. Selectivedecline of Nogo mRNA in the aging brain. Neuroreport 17, 913–916.

van Haaren, F., Meyer, M.E., 1991. Sex differences in locomotor activity after acuteand chronic cocaine administration. Pharmacology Biochemistry and Behavior39, 923–927.

Wiley, J.L., Evans, R.L., Grainger, D.B., Nicholson, K.L., 2008. Age-dependentdifferences in sensitivity and sensitization to cannabinoids and ‘club drugs’ inmale adolescent and adult rats. Addiction Biology 13, 277–286.

Woods, R.P., Grafton, S.T., Holmes, C.J., Cherry, S.R., Mazziotta, J.C., 1998a. Auto-mated image registration: I. General methods and intrasubject, intramodalityvalidation. Journal of Computer Assisted Tomography 22, 139–152.

Woods, R.P., Grafton, S.T., Watson, J.D., Sicotte, N.L., Mazziotta, J.C., 1998b. Auto-mated image registration: II. Intersubject validation of linear and nonlinearmodels. Journal of Computer Assisted Tomography 22, 153–165.

Yang, H.J., Wang, L., Cheng, Q., Xu, H., 2011. Abnormal behaviors and microstruc-tural changes in white matter of juvenile mice repeatedly exposed to amphe-tamine. Schizophrenia Research and Treatment 2011, 542896.

P.A. Narayana et al. / Psychiatry Research: Neuroimaging 221 (2014) 220–230230