Advertisement

Myelin and Axon Abnormalities in Schizophrenia Measured with Magnetic Resonance Imaging Techniques

      Background

      In schizophrenia (SZ), disturbances in integration of activity among brain regions seem to be as important as abnormal activity of any single region. Brain regions are connected through white matter (WM) tracts, and diffusion tensor imaging has provided compelling evidence for WM abnormalities in SZ. However, diffusion tensor imaging alone cannot currently pinpoint the biological basis of these abnormalities.

      Methods

      In this study, we combined a myelin-specific and an axon-specific magnetic resonance imaging approach to examine potentially distinct abnormalities of WM components in SZ. Magnetization transfer ratio (MTR) provides information on myelin content, whereas diffusion tensor spectroscopy provides information on metabolite diffusion within axons. We collected data from a 1×3×3 cm voxel within the right prefrontal cortex WM at 4 Tesla and studied 23 patients with SZ and 22 age- and sex-matched healthy control participants.

      Results

      The MTR was significantly reduced in SZ, suggesting reduced myelin content. By contrast, the apparent diffusion coefficient of N-acetylaspartate (NAA) was significantly elevated, suggesting intra-axonal abnormalities. Greater abnormality of both MTR and the apparent diffusion coefficient of NAA correlated with more adverse outcomes in the patient group.

      Conclusions

      The results suggest that WM abnormalities in SZ include both abnormal myelination and abnormal NAA diffusion within axons. These processes might be associated with abnormal signal transduction and abnormal information processing in SZ.

      Key Words

      Diffusion tensor imaging (DTI) provides information about water molecule diffusion and yields three diffusion eigenvalues labeled λ1, λ2, and λ3 from largest to smallest. The white matter (WM) of the brain contains axon fibers, and water molecular diffusion takes place along the long axis of these fibers (axial diffusivity [AD] = λ1) more than perpendicular to it (radial diffusivity [RD] = [λ2 + λ3]/2). Fractional anisotropy (FA) reflects directionality of diffusion (isotropic vs. anisotropic). Finally, mean or apparent diffusion coefficient (ADC) (ADC = [λ1 + λ2 + λ3]/3) reflects the distance traveled by a molecule in unit time, partly reflecting geometry of the surrounding space. Past DTI studies have provided strong evidence for widespread disruptions in WM integrity in schizophrenia (SZ). The FA reductions are associated with passivity phenomena (
      • Sim K.
      • Yang G.L.
      • Loh D.
      • Poon L.Y.
      • Sitoh Y.Y.
      • Verma S.
      • et al.
      White matter abnormalities and neurocognitive deficits associated with the passivity phenomenon in schizophrenia: A diffusion tensor imaging study.
      ), auditory hallucinations (
      • Shergill S.S.
      • Kanaan R.A.
      • Chitnis X.A.
      • O’Daly O.
      • Jones D.K.
      • Frangou S.
      • et al.
      A diffusion tensor imaging study of fasciculi in schizophrenia.
      ), impairments in working memory (
      • Karlsgodt K.H.
      • van Erp T.G.
      • Poldrack R.A.
      • Bearden C.E.
      • Nuechterlein K.H.
      • Cannon T.D.
      Diffusion tensor imaging of the superior longitudinal fasciculus and working memory in recent-onset schizophrenia.
      ,
      • Kubicki M.
      • Westin C.F.
      • Nestor P.G.
      • Wible C.G.
      • Frumin M.
      • Maier S.E.
      • et al.
      Cingulate fasciculus integrity disruption in schizophrenia: A magnetic resonance diffusion tensor imaging study.
      ) and executive function (
      • Rusch N.
      • Spoletini I.
      • Wilke M.
      • Bria P.
      • Di Paola M.
      • Di Iulio F.
      • et al.
      Prefrontal-thalamic-cerebellar gray matter networks and executive functioning in schizophrenia.
      ,
      • Takei K.
      • Yamasue H.
      • Abe O.
      • Yamada H.
      • Inoue H.
      • Suga M.
      • et al.
      Structural disruption of the dorsal cingulum bundle is associated with impaired Stroop performance in patients with schizophrenia.
      ,
      • Nestor P.G.
      • Kubicki M.
      • Gurrera R.J.
      • Niznikiewicz M.
      • Frumin M.
      • McCarley R.W.
      • et al.
      Neuropsychological correlates of diffusion tensor imaging in schizophrenia.
      ,
      • Nestor P.G.
      • Kubicki M.
      • Niznikiewicz M.
      • Gurrera R.J.
      • McCarley R.W.
      • Shenton M.E.
      Neuropsychological disturbance in schizophrenia: A diffusion tensor imaging study.
      ), and abnormal functional magnetic resonance imaging (MRI) connectivity (
      • Spoletini I.
      • Cherubini A.
      • Di Paola M.
      • Banfi G.
      • Rusch N.
      • Martinotti G.
      • et al.
      Reduced fronto-temporal connectivity is associated with frontal gray matter density reduction and neuropsychological deficit in schizophrenia.
      ,
      • Schlosser R.G.
      • Nenadic I.
      • Wagner G.
      • Gullmar D.
      • von Consbruch K.
      • Kohler S.
      • et al.
      White matter abnormalities and brain activation in schizophrenia: A combined DTI and fMRI study.
      ). A related literature provides evidence of deficits in integration of large-scale neuronal networks (
      • Williamson P.
      Are anticorrelated networks in the brain relevant to schizophrenia?.
      ,
      • Garrity A.G.
      • Pearlson G.D.
      • McKiernan K.
      • Lloyd D.
      • Kiehl K.A.
      • Calhoun V.D.
      Aberrant “default mode” functional connectivity in schizophrenia.
      ,
      • Whitfield-Gabrieli S.
      • Thermenos H.W.
      • Milanovic S.
      • Tsuang M.T.
      • Faraone S.V.
      • McCarley R.W.
      • et al.
      Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia.
      ) and in expression of myelin- and oligodendrocyte-related genes postmortem in SZ (
      • Tkachev D.
      • Mimmack M.L.
      • Ryan M.M.
      • Wayland M.
      • Freeman T.
      • Jones P.B.
      • et al.
      Oligodendrocyte dysfunction in schizophrenia and bipolar disorder.
      ). Thus, abnormal integration of activity across brain regions seems critical to SZ pathophysiology.
      Although WM abnormalities are central to SZ as an abnormal connection syndrome (
      • Stephan K.E.
      • Friston K.J.
      • Frith C.D.
      Dysconnection in schizophrenia: From abnormal synaptic plasticity to failures of self-monitoring.
      ,
      • Paus T.
      • Keshavan M.
      • Giedd J.N.
      Why do many psychiatric disorders emerge during adolescence?.
      ,
      • Kubicki M.
      • McCarley R.
      • Westin C.F.
      • Park H.J.
      • Maier S.
      • Kikinis R.
      • et al.
      A review of diffusion tensor imaging studies in schizophrenia.
      ), the link between DTI and brain function remains abstract, because of the nonspecific nature of the DTI signal (
      • Whitford T.J.
      • Ford J.M.
      • Mathalon D.H.
      • Kubicki M.
      • Shenton M.E.
      Schizophrenia, myelination, and delayed corollary discharges: A hypothesis.
      ). The FA, AD, and RD abnormalities are commonly interpreted as reflecting loss of “white matter integrity,” but its exact nature cannot be determined with DTI alone. Water exists in intra- and extracellular compartments, and there is exchange of water molecules between the two. Thus DTI abnormalities might reflect multiple processes (demyelination, fiber crossing, axonal swelling, or atrophy) and even different abnormalities in different cases (
      • Alexander A.L.
      • Lee J.E.
      • Lazar M.
      • Field A.S.
      Diffusion tensor imaging of the brain.
      ).
      Separate in vivo measures of axon and myelin integrity would be valuable to address this issue. Notably, indices of axonal diameter and myelin sheath thickness would allow one to predict whether signal conduction speed is abnormal in SZ brains. Here, we use two MR-based approaches to probe specific WM abnormalities in SZ: magnetization transfer ratio (MTR); and diffusion tensor spectroscopy (DTS). The MTR relies on magnetization exchange between water molecules in different physical environments. In biological tissue, “bound” water molecules around myelin lipids exchange protons with “free” water molecules. This exchange can be measured with a magnetization transfer paradigm where signal from “bound water” is saturated and the loss of “free water” signal (reflecting transfer to “bound water”) is measured. The larger the WM myelin component, the greater is the proton exchange, and the higher the MTR. The MTR is reduced in SZ, suggesting reduced myelin complement in this condition (
      • Kubicki M.
      • Park H.
      • Westin C.F.
      • Nestor P.G.
      • Mulkern R.V.
      • Maier S.E.
      • et al.
      DTI and MTR abnormalities in schizophrenia: analysis of white matter integrity.
      ,
      • Price G.
      • Cercignani M.
      • Chu E.M.
      • Barnes T.R.
      • Barker G.J.
      • Joyce E.M.
      • et al.
      Brain pathology in first-episode psychosis: magnetization transfer imaging provides additional information to MRI measurements of volume loss.
      ), although a recent study reported partially discrepant results (
      • Mandl R.C.
      • Schnack H.G.
      • Luigjes J.
      • van den Heuvel M.P.
      • Cahn W.
      • Kahn R.S.
      • Hulshoff Pol H.E.
      Tract-based analysis of magnetization transfer ratio and diffusion tensor imaging of the frontal and frontotemporal connections in schizophrenia.
      ).
      The DTS measures the diffusion of intracellular metabolites such as N-acetylaspartate (NAA). Because NAA is located exclusively in neurons and almost exclusively in the cytosol where diffusion is less restricted than within organelles (
      • Tsai G.
      • Coyle J.T.
      N-acetylaspartate in neuropsychiatric disorders.
      ), NAA diffusion provides specific information about intra-neuronal structure. The DTS measures are based on molecular Brownian motion and are independent of metabolite concentration and transverse (T2) relaxation times; therefore the NAA reductions (
      • Lim K.O.
      • Adalsteinsson E.
      • Spielman D.
      • Sullivan E.V.
      • Rosenbloom M.J.
      • Pfefferbaum A.
      Proton magnetic resonance spectroscopic imaging of cortical gray and white matter in schizophrenia.
      ) and NAA T2 abnormalities (
      • Ongur D.
      • Prescot A.P.
      • Jensen J.E.
      • Rouse E.D.
      • Cohen B.M.
      • Renshaw P.F.
      • et al.
      T2 relaxation time abnormalities in bipolar disorder and schizophrenia.
      ) observed in SZ do not confound NAA diffusion. The DTS approaches have been validated in a variety of contexts, including as probes of cellular diffusion (
      • Ackerman J.J.
      • Neil J.J.
      The use of MR-detectable reporter molecules and ions to evaluate diffusion in normal and ischemic brain.
      ) and in seminal studies of axon diameter (
      • Upadhyay J.
      • Hallock K.
      • Ducros M.
      • Kim D.S.
      • Ronen I.
      Diffusion tensor spectroscopy and imaging of the arcuate fasciculus.
      ). The only clinical studies using DTS are in acute cerebral ischemia, where NAA ADC is significantly reduced (
      • Harada M.
      • Uno M.
      • Hong F.
      • Hisaoka S.
      • Nishitani H.
      • Matsuda T.
      Diffusion-weighted in vivo localized proton MR spectroscopy of human cerebral ischemia and tumor.
      ,
      • Abe O.
      • Okubo T.
      • Hayashi N.
      • Saito N.
      • Iriguchi N.
      • Shirouzu I.
      • et al.
      Temporal changes of the apparent diffusion coefficients of water and metabolites in rats with hemispheric infarction: Experimental study of transhemispheric diaschisis in the contralateral hemisphere at 7 tesla.
      ,
      • Dreher W.
      • Busch E.
      • Leibfritz D.
      Changes in apparent diffusion coefficients of metabolites in rat brain after middle cerebral artery occlusion measured by proton magnetic resonance spectroscopy.
      ), and MELAS (mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes), where it is elevated (
      • Liu Z.
      • Liu X.
      • Hui L.
      • Zhao D.
      • Wang X.
      • Xie S.
      • et al.
      The appearance of ADCs in the non-affected areas of the patients with MELAS.
      ). The DTS parameter of greatest interest in the present study is NAA ADC, which has no directionality. By contrast, FA as well as AD and RD scale with directionality of structures in the voxel, and macroscopic curvature artifacts render them uninterpretable in a large voxel (see Discussion). The NAA ADC is informative about axon abnormalities: demyelination with preserved axon diameter would leave NAA ADC normal, whereas changes in NAA diffusion within axons with preserved myelination would modify NAA ADC.
      The NAA ADC can be determined by axonal geometry or by NAA distribution within axonal organelles (e.g., mitochondria). Interactions between axon health and myelin sheath thickness are complex and bidirectional. Larger axons have thicker myelin sheaths, and vice versa. The ratio between axon diameter and fiber diameter (defined as axon diameter + myelin sheath thickness) is termed the “g-ratio.” The g-ratio evolves during brain development and reaches a level of .6 in adulthood (
      • Chomiak T.
      • Hu B.
      What is the optimal value of the g-ratio for myelinated fibers in the rat CNS? A theoretical approach.
      ,
      • Kandel E.R.
      • Schwartz J.H.
      • Jessell T.M.
      Principles of Neural Science.
      ,
      • Paus T.
      • Toro R.
      Could sex differences in white matter be explained by g ratio?.
      ). Divergence from this optimal g-ratio in either direction is associated with abnormalities in conduction speed (
      • Nave K.A.
      Myelination and the trophic support of long axons.
      ).
      The combination of MTR and DTS affords the ability to probe axon versus myelin-related abnormalities separately in the human WM. We hypothesized, on the basis of strong evidence for myelination abnormalities (
      • Uranova N.A.
      • Vostrikov V.M.
      • Vikhreva O.V.
      • Zimina I.S.
      • Kolomeets N.S.
      • Orlovskaya D.D.
      The role of oligodendrocyte pathology in schizophrenia.
      ,
      • Flynn S.W.
      • Lang D.J.
      • Mackay A.L.
      • Goghari V.
      • Vavasour I.M.
      • Whittall K.P.
      • et al.
      Abnormalities of myelination in schizophrenia detected in vivo with MRI, and post-mortem with analysis of oligodendrocyte proteins.
      ,
      • McCullumsmith R.E.
      • Gupta D.
      • Beneyto M.
      • Kreger E.
      • Haroutunian V.
      • Davis K.L.
      • et al.
      Expression of transcripts for myelination-related genes in the anterior cingulate cortex in schizophrenia.
      ,
      • Hakak Y.
      • Walker J.R.
      • Li C.
      • Wong W.H.
      • Davis K.L.
      • Buxbaum J.D.
      • et al.
      Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia.
      ) as well as a mechanistic relationship between developmental myelination and SZ (
      • Budel S.
      • Padukkavidana T.
      • Liu B.P.
      • Feng Z.
      • Hu F.
      • Johnson S.
      • et al.
      Genetic variants of Nogo-66 receptor with possible association to schizophrenia block myelin inhibition of axon growth.
      ,
      • Bernstein H.G.
      • Steiner J.
      • Bogerts B.
      Glial cells in schizophrenia: Pathophysiological significance and possible consequences for therapy.
      ), that myelin sheath thickness and MTR are reduced in SZ. Given the paucity of published information on axon structure in SZ, we could not predict NAA ADC changes.

      Methods and Materials

      Participants

      After approval by the McLean Hospital Institutional Review Board, we recruited 22 healthy control subjects from the community and 23 participants with SZ from the clinical services at McLean Hospital. Demographic and clinical characteristics of the study participants are provided in Table 1. See Supplement 1 for details of inclusion/exclusion criteria, participant screening, and standard study procedures.
      Table 1Demographic and Clinical Characteristics of Study Participants
      Healthy Control (n = 22)Schizophrenia (n = 23)Statistical Evaluation
      Age (yrs)31.4±6.734.0±9.2F43,1 = 1.19; p = .286
      Gender12 M, 10 F14 M, 9 Fχ2 = .18; p = .668
      BMI23.4±2.826.9±4.5F42,1 = 9.13; p = .004
      Education
      Education code: 3: graduated high school; 4: part college; 5: graduated 2-year college; 6: graduated 4-year college; 7: part graduate/professional school; 8: completed graduate/professional school.
       6.8±1.0 4.4±1.4F42,1 = 42.70; p<.001
      Parental SES
      Parental SES calculated according to the Hollingshead scale.
       5.9±1.0 5.5±1.6F26,1 = .46; p = .505
      Age at Onset (yrs)21.8±6.8
      Lifetime Number of Suicide Attempts 1.5±2.5
      Lifetime Number of Hospital Stays 6.7±6.1
      MADRS 10.2±10.4
      YMRS 8.1±7.8
      PANSS 49.8±14.0
      Lithium3
      Anticonvulsants4
      SGAs19
      FGAs2
      CPZ Equivalents 501±459
      Benzodiazepines7
      Values given are number, unless otherwise indicated. BMI, body mass index; CPZ, chlorpromazine; FGA, first-generation antipsychotic; MADRS, Montgomery–Åsberg Depression Rating Scale; PANSS, Positive and Negative Syndrome Scale; SES, socioeconomic status; SGA, second-generation antipsychotic; YMRS, Young Mania Rating Scale.
      a Education code: 3: graduated high school; 4: part college; 5: graduated 2-year college; 6: graduated 4-year college; 7: part graduate/professional school; 8: completed graduate/professional school.
      b Parental SES calculated according to the Hollingshead scale.

      MRI and Spectroscopy

      See Supplement 1 for details of anatomic imaging and voxel placement (Figure 1).
      Figure thumbnail gr1
      Figure 1Representative axial images depicting the location of our 1×3×3 cm white matter voxel in the right prefrontal cortex.
      MTR. The MTR experiment relies on measuring water signal magnitude in the presence and absence of a saturation pulse, which causes saturation of signal coming from “bound” water molecules. Because there is exchange between “bound” and “free” water molecules, the saturation pulse measurably attenuates the signal coming from “free” water molecules (measured as the water resonance after the pulse) (Figure 2A). The MTR is calculated on the basis of water signal intensity acquired in the presence (Ms) and absence (Mc) of the saturation pulse (MTR = [Mc−Ms]/Mc).
      Figure thumbnail gr2
      Figure 2(A) Magnetic resonance spectra of the water resonance acquired from a healthy control participant in the presence (red) and absence (black) of the saturation pulse. The reduction in magnetization of the water resonance (ΔM) and the “restricted water” resonance (amplified scale) are also shown. The ΔM forms the basis of calculations for the magnetization transfer ratio. (B) Magnetization transfer ratio presented from healthy control subjects (green), schizophrenia patients (red), and from a phantom (blue). The region highlighted by the box exhibits direct radiofrequency off-resonance effects and hence was excluded from the calculations presented in this paper.
      We used a BISTRO saturation pulse train (
      • de Graaf R.A.
      • Luo Y.
      • Garwood M.
      • Nicolay K.
      B1-insensitive, single-shot localization and water suppression.
      ) constructed with multiple hyperbolic Sec pulses (width = 50 msec) with varied radiofrequency (RF) pulse amplitudes and applied at the beginning of a standard point-resolved spectroscopy (PRESS) sequence (before the 90-degree pulse) to saturate “bound-water” signal with a specific frequency offset (
      • de Graaf R.A.
      • Luo Y.
      • Garwood M.
      • Nicolay K.
      B1-insensitive, single-shot localization and water suppression.
      ,

      Chen W, Luo Y, Merkle H, Zhu X-H, Adriany G, Garwood M, et al. (1995): Low-power, B1-insensitive, frequency-selective saturation pulse for use of water suppression and saturation transfer in NMR. SMR, 3rd Annual Meeting. Nice, France: Society for Magnetic Resonance, 1016.

      ). Data were obtained in 50-Hz steps at a range of frequencies offset 400–1000 Hz in either direction from the water signal, and a single MTR number was calculated by averaging across frequencies. Saturation time (tsat) was 2.6 sec with repetition time/echo time = 3000/30 msec and repetitions = 2.
      DTS Measurements. The standard PRESS sequence was modified by incorporating diffusion gradients for DTS measurements. Bipolar diffusion gradients with six directions—(1,1,0) (1,0,1) (0,1,1) (−1,1,0) (−1,0,1) (0,−1,1)—and one control (0,0,0) (totaling seven spectra) were applied to calculate diffusion tensors of signal from water and metabolites. The applied b value was 1412 sec/mm2, calibrated by a homemade phantom with water ADC assumed to be 2.1×10−3 mm2/sec at room temperature (approximately 20°C) (
      • Kan H.E.
      • Techawiboonwong A.
      • van Osch M.J.
      • Versluis M.J.
      • Deelchand D.K.
      • Henry P.G.
      • et al.
      Differences in apparent diffusion coefficients of brain metabolites between grey and white matter in the human brain measured at 7 T.
      ). In these measurements, repetition time/echo time = 3000/135 msec, diffusion time (Dt) = 60 msec, repetitions = 96 and 4 for metabolites and water diffusion measurements, respectively. Metabolite spectra were acquired with water saturation with VAPOR (
      • Tkac I.
      • Starcuk Z.
      • Choi I.Y.
      • Gruetter R.
      In vivo 1H NMR spectroscopy of rat brain at 1 ms echo time.
      ). Free induction decays were stored separately before averaging for correction of frequency- and phase-drifts and eddy currents resulting from diffusion gradients or instability of machine hardware. Total experiment time including MTR and DTS measurements of water and metabolites was approximately 70 min. The water data from the DTS experiment are analogous to DTI data from the literature, with one major exception: they were collected from a large 1×3×3 cm3 voxel. In addition, before or after each human subject study, we carried out a phantom DTS scan to correct for measurement errors from potential machine instability.

      MRI and Magnetic Resonance Spectroscopy Data Processing/Analysis

      An MR physicist (F.D.) processed all MRI/magnetic resonance spectroscopy (MRS) data blind to diagnosis. Post-processing of the free induction decays—including apodization, Fourier transformation, frequency, phase, and eddy current correction of individual spectra in the DTS experiment—as well as calculation of MTR and DTS constants were carried out with software provided in the Varian Console and home-grown software running on MATLAB (MathWorks, Natick, Massachusetts). Note that MTR and DTS measurements depend on relative signal change with saturation RF pulse or diffusion gradient, respectively. We digitized the water or NAA signal (resonance peak area) and normalized it to baseline (i.e., to the signal without RF saturation in MTR or to that without diffusion gradients in DTS). The units for ADC, RD, and AD are mm2/sec×10−3.
      We collected creatine (Cr) and choline (Cho) data along with NAA in our DTS studies. The signal-to-noise ratio (SNR) is lower for these metabolites than for NAA, but it was possible to carry out analyses of Cho in our DTS data. This is valuable, because Cho is compartmentalized differently from NAA (i.e., more of it is found in astrocytes than neurons) (
      • Choi J.K.
      • Dedeoglu A.
      • Jenkins B.G.
      Application of MRS to mouse models of neurodegenerative illness.
      ). If Cho and NAA ADC show differential patterns in SZ, this would support the neuron-selective significance of NAA ADC.

      Statistical Approach

      All analyses were carried out with SPSS (version 18; SPSS, Chicago, IL). The statistical plan had three stages: tests of data quality; tests of our main hypotheses; and exploratory tests of associations between multiple variables in the dataset. First, two-sample t tests and χ2 tests compared sample characteristics and SNR for the 135-msec DTS-PRESS spectrum (MRS data quality measure) across groups.
      Second, we entertained our two main hypotheses: 1) that there would be a reduction in MTR; and 2) NAA ADC might be abnormal in schizophrenia. These hypotheses were tested with two general linear models with MTR and NAA ADC as outcomes and diagnosis as predictor. Given the richness of the data we collected, we carried out parallel secondary analyses on NAA RD, AD and FA, and water RD, AD, and FA. Because age and smoking can impact WM health, we reran the main analyses with these variables as covariates.
      Third, we carried out a series of correlation analyses with Pearson’s coefficients (or Spearman where specifically mentioned for variables with skewed distribution). We examined correlations between the various diffusion variables to detect any possible structured covariance. We also examined the relationship between MTR and NAA ADC with age, education level, and body mass index (BMI) for the full dataset and MTR and NAA ADC with duration of illness, lifetime number of suicide attempts, lifetime number of hospital stays, North American Adult Reading Test score, Multnomah Community Ability Scale score, chlorpromazine equivalents, Positive and Negative Syndrome Scale, Young Mania Rating Scale, and Montgomery–Åsberg Depression Rating Scale scores for the SZ group. In addition, we carried out analyses of variance with sex and race as independent variables and MTR or NAA ADC as the dependent variable. We did not control for multiple comparisons in any of these exploratory analyses, because our goal was to allow detection of even modest relationships so they could be pursued in future studies. We were willing to accept the risk of Type I error inherent in this approach, because these did not concern our primary hypotheses and we did not have adequate power in this small clinical sample to correct for multiple comparisons.

      Results

      See Table 1 for demographic and clinical variables; the two groups were well-matched with the exceptions usually noted in samples of patients with schizophrenia: BMI; and participant educational attainment. To assess the reliability of our measures, we first carried out a test-retest study (Supplement 1).

      MTR Spectroscopy

      The MTR measurements are described in Figure 2. Data from a phantom aqueous sodium chloride solution showed a very low MTR of<1%, a face-valid finding because there is no “bound” water in an aqueous solution. In the human brain, “bound” water molecules (i.e., those interacting with lipids and proteins) cause a loss of water signal intensity, leading to a non-zero MTR. There was a significant reduction in MTR in SZ as compared with healthy control subjects [F36,1 = 5.339, p = .027], and this remained when age and smoking were added as covariates [F36,1 = 5.682, p = .023] (Table 2).
      Table 2MTR and DTS Data Summary
      Normal ControlSchizophreniaStatistical EvaluationEffect Size (Cohen’s d)
      MTR.17±.02.15±.03F36,1 = 5.339, p = .027.78
      NAA RD.15±.04.17±.05F41,1 = 1.674, p = .203.44
      NAA AD.33±.09.40±.10F41,1 = 4.189, p = .047
      p< .05.
      .74
      NAA ADC.21±.05.25±.05F41,1 = 6.348, p = .016
      p< .05.
      .80
      NAA FA.48±.16.52±.16F41,1 = .696, p = .409.25
      Water RD.53±.06.59±.10F41,1 = 4.865, p = .033
      p< .05.
      .73
      Water AD.76±.09.87±.14F41,1 = 8.417, p = .006
      p< .05.
      .93
      Water ADC.61±.05.68±.11F41,1 = 7.687, p = .008
      p< .05.
      .82
      Water FA.24±.11.25±.07F41,1 = .422, p = .519.11
      Statistical analyses are described in the text. AD, axial diffusivity; ADC, apparent diffusion coefficient; DTS, diffusion tensor spectroscopy; FA, fractional anisotropy; MTR, magnetization transfer ratio; NAA, N-acetylaspartate; RD, radial diffusivity.
      a p< .05.

      DTS

      The DTS measurements in a phantom preparation ([NAA] = 12 mmol/L, [Cr] = 8 mmol/L, and [Cho] = 3 mmol/L, pH = 7.0) demonstrated isotropic diffusion, as expected in a structure-free medium (FA≤.08 for all three chemicals), and ADCs were in good agreement with published values (.64±.08, .74±.06, and .92±.06 for NAA, Cr, and Cho, respectively) (
      • Kan H.E.
      • Techawiboonwong A.
      • van Osch M.J.
      • Versluis M.J.
      • Deelchand D.K.
      • Henry P.G.
      • et al.
      Differences in apparent diffusion coefficients of brain metabolites between grey and white matter in the human brain measured at 7 T.
      ). The spectra obtained during a typical in vivo DTS experiment are shown in Figure 3. Note that we show water-suppressed spectra for simplicity, although we also collected water-unsuppressed spectra for calculation of water diffusion. See Supplement 1 for discussion of SNR in our DTS data.
      Figure thumbnail gr3
      Figure 3Sample water-suppressed magnetic resonance spectroscopy spectra showing the modulation of metabolite signal with diffusion gradients. The top row shows data from a typical healthy control subject; the bottom row shows data from a schizophrenia patient. In each row, the leftmost spectrum is with no gradients applied, and the next six spectra show a variety of x/y/z gradients as shown. Note the variable decrement in metabolite signal with differing gradients, giving rise to the calculation of the diffusion tensor. Cho, choline; Cr, creatine; NAA, N-acetylaspartate.
      The NAA ADC was significantly elevated in SZ when compared with control subjects [F41,1 = 6.348, p = .016], and this was true when age and smoking are added to the model [F41,1 = 5.500, p = .023]. This measure was not correlated with MTR (R =−.242, p = .156). Water ADC was also elevated in SZ [F41,1 = 7.687, p = .008] but not correlated with MTR (R = −.129, p = .447). For reasons discussed in the following text, we did not consider FA a primary outcome (NAA FA and water FA in SZ were not significantly different from control in this study; see Table 2 for details). In addition, Cho ADC was not significantly different between healthy control and SZ groups (.20±.06 and .22±.05, respectively) [F37,1 = .227, p = .637].
      As expected, there were numerous statistically significant correlations (not shown) among the NAA diffusion parameters as well as among the water diffusion parameters (exceptions were NAA RD-AD, NAA FA-ADC, and water FA-ADC). By contrast, there were no correlations between any water and any NAA diffusion parameters, suggesting these measures were at least partially independent, although they share some common mechanisms (e.g., intra-axonal water and NAA diffusion).

      MTR, DTS, and Demographic/Clinical Variables

      The MTR correlated negatively with BMI (R = −.449, p = .005) and positively with education level (R = .440, p = .006) in the full dataset. It also correlated with number of lifetime suicide attempts (Spearman R = −.553, p = .014) and number of lifetime hospital stays (R = −.620, p = .005) among patients (Figure S1 in Supplement 1). The NAA ADC correlated negatively with NAART score (R = −.462, p = .040) and education level (R = −.393, p = .010). By contrast, water ADC was not correlated with NAART score (R = .309, p = .173) or education level (R = −.141, p = .367). Because the two groups were not matched for BMI and education, we also ran the MTR/BMI (R = −.591, p = .008), MTR/education (R = −.118, p = .592), and NAA ADC/education (R = −.544, p = .011) correlations within the SZ group only. No other correlation analysis was statistically significant, despite the liberal approach of not correcting for multiple comparisons.

      Discussion

      We applied a combined MTR-DTS approach to probe microstructural WM abnormalities in chronically ill SZ patients. We implemented several data quality measures: a test-retest study in healthy control subjects, phantom calibration of each human acquisition, frequency/phase/eddy current correction of individual diffusion spectra, and SNR calculation for individual spectra. We found that MTR was reduced and NAA ADC was elevated in SZ, suggesting that WM pathology in SZ is driven by both myelination deficits and axon abnormalities.
      Several additional features of the DTS data lend face validity to our findings. For example, the water FA and ADC and the NAA FA and ADC values for healthy control subjects are similar to those observed in other DTI and DTS studies [e.g., (
      • Upadhyay J.
      • Hallock K.
      • Ducros M.
      • Kim D.S.
      • Ronen I.
      Diffusion tensor spectroscopy and imaging of the arcuate fasciculus.
      ,
      • Camchong J.
      • Lim K.O.
      • Sponheim S.R.
      • Macdonald A.W.
      Frontal white matter integrity as an endophenotype for schizophrenia: Diffusion tensor imaging in monozygotic twins and patients’ nonpsychotic relatives.
      )]. The NAA is a larger molecule, diffuses more slowly than water, and is predicted to have a lower AD; this is exactly what we observe. In addition, because NAA is intracellular, it is predicted to have more anisotropic diffusion than water, and we find that NAA FA is higher than water FA in healthy control subjects. However, note that FA is sensitive to noise, and the NAA resonance has lower signal-to-noise than that of water. Finally, our finding of reduced MTR is consistent with the elevated water RD in SZ, because the latter is proposed as an indicator of myelin reduction.
      The clinical significance of these findings was highlighted because low MTR was associated with worse educational attainment, markers of a more severe phenotype (lifetime hospital stays and suicide attempts), and higher BMI, whereas higher NAA ADC was associated with lower NAART scores and educational attainment. Thus, for each measure, the direction of change in patients compared with control subjects (reduced MTR, elevated NAA ADC) was also associated with more adverse outcomes within the patient group. In addition, we did not see similar correlations with water ADC, suggesting that there is additional value in measuring NAA ADC. This suggests the WM abnormalities we observed in schizophrenia might be related to factors that determine functional outcomes in this condition. Although we cannot currently propose a mechanism for these effects, future studies can examine the relationship between specific cognitive functions or clinical symptoms and MTR/DTS in specific WM tracts.
      Our findings suggest that both myelination (measured by MTR) and axonal abnormalities (measured by DTS) play a role in WM abnormalities in SZ. One interpretation consistent with our hypotheses is that myelin is reduced in SZ accompanied by an increase in intra-axonal space available for diffusion (increase in axonal diameter or reduced hindrance within axons). The axon-selectivity of the NAA ADC findings is further supported by the absence of similar changes in Cho ADC. The myelin and axon changes would act in concert to lead to abnormal signal transduction between brain regions in SZ. In addition, they would each reduce the anisotropy of water molecule diffusion, leading to well-documented DTI abnormalities. Consistent with this assumption, we also observe changes in water diffusion along with those in MTR and NAA ADC. The simultaneous elevation of NAA and water ADC in SZ is distinct from stroke, where both measures are reduced (
      • Dreher W.
      • Busch E.
      • Leibfritz D.
      Changes in apparent diffusion coefficients of metabolites in rat brain after middle cerebral artery occlusion measured by proton magnetic resonance spectroscopy.
      ). By contrast, it is noteworthy that the MTR and DTS measures did not correlate with each other. This suggests the possibility of independent mechanisms leading to myelin and axon abnormalities. Future longitudinal studies in early stages of schizophrenia might be instructive in how these mechanisms evolve.
      This interpretation suggests abnormalities in the g-ratio in schizophrenia. We cannot calculate a g-ratio from the current dataset, because MTR and DTS data have different units. Therefore, we calculated NAA ADC/MTR as a simpler index. This ratio is 1.35±.27 for control subjects and 1.66±.39 for patients (mean±SD), a 24% between-group difference (p = .007). We note that this difference, although not a primary outcome of this study, is substantial and highly significant.
      Although abnormalities in axonal geometry are one possible explanation for our DTS data, others are possible. For example, molecular diffusion properties can be affected by membrane permeability during anesthesia (
      • Valette J.
      • Guillermier M.
      • Besret L.
      • Hantraye P.
      • Bloch G.
      • Lebon V.
      Isoflurane strongly affects the diffusion of intracellular metabolites, as shown by 1H nuclear magnetic resonance spectroscopy of the monkey brain.
      ), and there might be membrane permeability abnormalities in SZ. In addition, abnormalities in the proportion of NAA localized in mitochondria or in NAA cleavage in extracellular space as part of myelin synthesis (
      • Chakraborty G.
      • Mekala P.
      • Yahya D.
      • Wu G.
      • Ledeen R.W.
      Intraneuronal N-acetylaspartate supplies acetyl groups for myelin lipid synthesis: Evidence for myelin-associated aspartoacylase.
      ) might also impact NAA ADC. Because mitochondrial abnormalities are reported in schizophrenia (
      • Ben-Shachar D.
      • Laifenfeld D.
      Mitochondria, synaptic plasticity, and schizophrenia.
      ), abnormal NAA distribution in mitochondria is an attractive alternative hypothesis to be pursued.
      The relationship between NAA diffusion and transverse T2 relaxation is intriguing, because each measure reflects a related but distinct aspect of the microenvironment of NAA. Although T2 relaxation is determined by spin-spin interactions between the index molecule and its immediate neighbors, diffusion reflects the distance traveled by a molecule in unit time. We recently reported, in a dataset partially overlapping with the current one, that NAA T2 relaxation time is shortened in SZ, whereas that of water is prolonged (
      • Du F.
      • Cooper A.
      • Cohen B.M.
      • Renshaw P.F.
      • Ongur D.
      Water and metabolite transverse T2 relaxation time abnormalities in the white matter in schizophrenia.
      ). This contrasts with the current findings of elevated NAA and water ADC in SZ. This pattern suggests that WM microenvironment changes in SZ might be more complex than only axonal geometry changes and involve both greater NAA diffusion and more frequent interaction with other molecules. In addition, there is no significant correlation between NAA ADC and NAA T2 in our data (R =−.214; p = .218), suggesting these measures reflect independent processes. Deeper insight into this issue would come from analysis of NAA T2 relaxation for multi-exponential decay. Multiple-component T2 relaxation might offer a clue that NAA signal arises from molecules in different environments. Here, we calculated T2 times on the basis of 4 echo times, not enough to explore multi-exponential decay. Future studies with more detailed T2 relaxation data are needed.
      Although DTS capitalizes on NAA diffusion in the same manner DTI does on water diffusion, brain NAA concentration is 5000-fold lower than brain water concentration (approximately 10 mmol/L and 50 mol/L, respectively). Thus, although DTI can achieve whole-brain coverage with millimetric voxels in minutes, we collect data from a single large voxel over many repetitions to obtain reliable DTS data. The large DTS voxel can be associated with macroscopic curvature effects: as axons course through the voxel, they might curve, and the DTS signal averaged over a large volume partly reflects curvature as opposed to diffusion. To address this confound, the results of a DTS study can be examined for macroscopic curvature effects. If macroscopic curvature effects are operative, RD and AD should covary (e.g., curved fibers would yield high RD and low AD, whereas straight fibers would yield low RD and high AD). If RD and AD vary independently of one another (as is the case with NAA), this suggests the findings are not secondary to macroscopic curvature. In addition, macroscopic curvature affects water and NAA parameters equally, whereas biologically specific changes cause independent variation in these parameters. The absence of covariation between NAA and water parameters in our study is reassuring in this regard. Nonetheless, we focused on ADC as the DTS parameter least affected by macroscopic curvature. This is because ADC has no directionality, whereas all other DTS measures do.
      In addition to these conceptual caveats, our study has several limitations. First is the potential for variable voxel placement, which could result in the inclusion of axons of different diameters in different brains, impacting the ADC measures. Our voxel is anchored by anatomical landmarks that improve reliability, as demonstrated by our test-retest study. Collecting data from the entire brain is currently possible for MTR but not for DTS. Chemical shift imaging can collect high-quality MRS data from the entire brain (
      • Maudsley A.A.
      • Domenig C.
      • Govind V.
      • Darkazanli A.
      • Studholme C.
      • Arheart K.
      • et al.
      Mapping of brain metabolite distributions by volumetric proton MR spectroscopic imaging (MRSI).
      ,
      • Posse S.
      • Otazo R.
      • Caprihan A.
      • Bustillo J.
      • Chen H.
      • Henry P.G.
      • et al.
      Proton echo-planar spectroscopic imaging of J-coupled resonances in human brain at 3 and 4 Tesla.
      ), but these are challenging to implement with diffusion gradients. Second, the NAA signal we measure contains contributions from NAA and N-acetylaspartylglutamate (NAAG). The NAAG is located both intra- and extracellularly (
      • Coyle J.T.
      The nagging question of the function of N-acetylaspartylglutamate.
      ), and our DTS measures might be confounded by this contamination. The NAAG is similar to NAA in chemical structure, so the two MRS signals are challenging to resolve. The NAAG concentration in the prefrontal cortex WM is 1.5 mmol/L in healthy individuals (of which an unknown fraction is extracellular) (
      • Pouwels P.J.
      • Frahm J.
      Differential distribution of NAA and NAAG in human brain as determined by quantitative localized proton MRS.
      ), whereas NAA concentrations are usually calculated at 10 mmol/L (
      • Govindaraju V.
      • Young K.
      • Maudsley A.A.
      Proton NMR chemical shifts and coupling constants for brain metabolites.
      ). Therefore, we do not expect NAAG contribution to be a major factor in our results. Third, MTR is not a specific measure of myelin content. The MTR abnormalities can arise from acute inflammation, edema, and other processes that impact brain water content (
      • Laule C.
      • Vavasour I.M.
      • Kolind S.H.
      • Li D.K.
      • Traboulsee T.L.
      • Moore G.R.
      • et al.
      Magnetic resonance imaging of myelin.
      ). This limits the utility of MTR in pathologies where gross brain water abnormalities are seen. There is no evidence for such abnormalities in SZ, and past applications of MTR in SZ have revealed subtle changes (
      • Kubicki M.
      • Park H.
      • Westin C.F.
      • Nestor P.G.
      • Mulkern R.V.
      • Maier S.E.
      • et al.
      DTI and MTR abnormalities in schizophrenia: analysis of white matter integrity.
      ,
      • Mandl R.C.
      • Schnack H.G.
      • Luigjes J.
      • van den Heuvel M.P.
      • Cahn W.
      • Kahn R.S.
      • Hulshoff Pol H.E.
      Tract-based analysis of magnetization transfer ratio and diffusion tensor imaging of the frontal and frontotemporal connections in schizophrenia.
      ). Others have used a “myelin water fraction” approach, which takes advantage of the differential T2 relaxation properties of water molecules trapped within myelin (
      • MacKay A
      • Laule C
      • Vavasour I
      • Bjarnason T
      • Kolind S
      • Madler B
      Insights into brain microstructure from the T2 distribution.
      ), but this approach is technically challenging. Fourth, the SZ patients in this study were chronically ill and taking medications. The DTI abnormalities are widely reported in chronically ill patients, in fact more consistently than in first-episode patients with SZ (
      • Friedman J.I.
      • Tang C.
      • Carpenter D.
      • Buchsbaum M.
      • Schmeidler J.
      • Flanagan L.
      • et al.
      Diffusion tensor imaging findings in first-episode and chronic schizophrenia patients.
      ) or those with other psychiatric diagnoses (

      Heng S, Song AW, Sim K (2010): White matter abnormalities in bipolar disorder: Insights from diffusion tensor imaging studies. J Neural Transm 117:639–654.

      ,
      • White T.
      • Nelson M.
      • Lim K.O.
      Diffusion tensor imaging in psychiatric disorders.
      ). In addition, we are not aware of any literature on psychotropic medications causing alterations in WM microstructure; there was no relationship between chlorpromazine equivalents and MTR or NAA ADC in our study. Nonetheless, we cannot rule out medication effects; this needs to be addressed in future studies. Finally, we did not correct for cardiac gating effects, which can impact the MRS signal. Because our MRS sequence involved interleaved acquisition, such effects were unlikely to impact the findings.
      In conclusion, we used a novel MRI/MRS approach to probe WM abnormalities in SZ and provided in vivo evidence for both abnormal axon geometry and myelination. Our findings suggest that signal transduction speeds are abnormal in SZ, possibly leading to information processing abnormalities and cognitive deficits. Future studies will focus on early stages of illness and on abnormalities in specific WM pathways with specific cognitive or clinical presentations.
      This work was funded by R01MH094594 (DO), R21MH092704 (FD), and the Shervert Frazier Research Institute at McLean Hospital to BMC.
      We are grateful to the participants for volunteering for research.
      The authors report no biomedical financial interests or potential conflicts of interest.

      Appendix A. Supplementary materials

      References

        • Sim K.
        • Yang G.L.
        • Loh D.
        • Poon L.Y.
        • Sitoh Y.Y.
        • Verma S.
        • et al.
        White matter abnormalities and neurocognitive deficits associated with the passivity phenomenon in schizophrenia: A diffusion tensor imaging study.
        Psychiatry Res. 2009; 172: 121-127
        • Shergill S.S.
        • Kanaan R.A.
        • Chitnis X.A.
        • O’Daly O.
        • Jones D.K.
        • Frangou S.
        • et al.
        A diffusion tensor imaging study of fasciculi in schizophrenia.
        Am J Psychiatry. 2007; 164: 467-473
        • Karlsgodt K.H.
        • van Erp T.G.
        • Poldrack R.A.
        • Bearden C.E.
        • Nuechterlein K.H.
        • Cannon T.D.
        Diffusion tensor imaging of the superior longitudinal fasciculus and working memory in recent-onset schizophrenia.
        Biol Psychiatry. 2008; 63: 512-518
        • Kubicki M.
        • Westin C.F.
        • Nestor P.G.
        • Wible C.G.
        • Frumin M.
        • Maier S.E.
        • et al.
        Cingulate fasciculus integrity disruption in schizophrenia: A magnetic resonance diffusion tensor imaging study.
        Biol Psychiatry. 2003; 54: 1171-1180
        • Rusch N.
        • Spoletini I.
        • Wilke M.
        • Bria P.
        • Di Paola M.
        • Di Iulio F.
        • et al.
        Prefrontal-thalamic-cerebellar gray matter networks and executive functioning in schizophrenia.
        Schizophr Res. 2007; 93: 79-89
        • Takei K.
        • Yamasue H.
        • Abe O.
        • Yamada H.
        • Inoue H.
        • Suga M.
        • et al.
        Structural disruption of the dorsal cingulum bundle is associated with impaired Stroop performance in patients with schizophrenia.
        Schizophr Res. 2009; 114: 119-127
        • Nestor P.G.
        • Kubicki M.
        • Gurrera R.J.
        • Niznikiewicz M.
        • Frumin M.
        • McCarley R.W.
        • et al.
        Neuropsychological correlates of diffusion tensor imaging in schizophrenia.
        Neuropsychology. 2004; 18: 629-637
        • Nestor P.G.
        • Kubicki M.
        • Niznikiewicz M.
        • Gurrera R.J.
        • McCarley R.W.
        • Shenton M.E.
        Neuropsychological disturbance in schizophrenia: A diffusion tensor imaging study.
        Neuropsychology. 2008; 22: 246-254
        • Spoletini I.
        • Cherubini A.
        • Di Paola M.
        • Banfi G.
        • Rusch N.
        • Martinotti G.
        • et al.
        Reduced fronto-temporal connectivity is associated with frontal gray matter density reduction and neuropsychological deficit in schizophrenia.
        Schizophr Res. 2009; 108: 57-68
        • Schlosser R.G.
        • Nenadic I.
        • Wagner G.
        • Gullmar D.
        • von Consbruch K.
        • Kohler S.
        • et al.
        White matter abnormalities and brain activation in schizophrenia: A combined DTI and fMRI study.
        Schizophr Res. 2007; 89: 1-11
        • Williamson P.
        Are anticorrelated networks in the brain relevant to schizophrenia?.
        Schizophr Bull. 2007; 33: 994-1003
        • Garrity A.G.
        • Pearlson G.D.
        • McKiernan K.
        • Lloyd D.
        • Kiehl K.A.
        • Calhoun V.D.
        Aberrant “default mode” functional connectivity in schizophrenia.
        Am J Psychiatry. 2007; 164: 450-457
        • Whitfield-Gabrieli S.
        • Thermenos H.W.
        • Milanovic S.
        • Tsuang M.T.
        • Faraone S.V.
        • McCarley R.W.
        • et al.
        Hyperactivity and hyperconnectivity of the default network in schizophrenia and in first-degree relatives of persons with schizophrenia.
        Proc Natl Acad Sci U S A. 2009; 106: 1279-1284
        • Tkachev D.
        • Mimmack M.L.
        • Ryan M.M.
        • Wayland M.
        • Freeman T.
        • Jones P.B.
        • et al.
        Oligodendrocyte dysfunction in schizophrenia and bipolar disorder.
        Lancet. 2003; 362: 798-805
        • Stephan K.E.
        • Friston K.J.
        • Frith C.D.
        Dysconnection in schizophrenia: From abnormal synaptic plasticity to failures of self-monitoring.
        Schizophr Bull. 2009; 35: 509-527
        • Paus T.
        • Keshavan M.
        • Giedd J.N.
        Why do many psychiatric disorders emerge during adolescence?.
        Nat Rev Neurosci. 2008; 9: 947-957
        • Kubicki M.
        • McCarley R.
        • Westin C.F.
        • Park H.J.
        • Maier S.
        • Kikinis R.
        • et al.
        A review of diffusion tensor imaging studies in schizophrenia.
        J Psychiatr Res. 2007; 41: 15-30
        • Whitford T.J.
        • Ford J.M.
        • Mathalon D.H.
        • Kubicki M.
        • Shenton M.E.
        Schizophrenia, myelination, and delayed corollary discharges: A hypothesis.
        Schizophr Bull. 2012; 38: 486-494
        • Alexander A.L.
        • Lee J.E.
        • Lazar M.
        • Field A.S.
        Diffusion tensor imaging of the brain.
        Neurotherapeutics. 2007; 4: 316-329
        • Kubicki M.
        • Park H.
        • Westin C.F.
        • Nestor P.G.
        • Mulkern R.V.
        • Maier S.E.
        • et al.
        DTI and MTR abnormalities in schizophrenia: analysis of white matter integrity.
        Neuroimage. 2005; 26: 1109-1118
        • Price G.
        • Cercignani M.
        • Chu E.M.
        • Barnes T.R.
        • Barker G.J.
        • Joyce E.M.
        • et al.
        Brain pathology in first-episode psychosis: magnetization transfer imaging provides additional information to MRI measurements of volume loss.
        Neuroimage. 2010; 49: 185-192
        • Mandl R.C.
        • Schnack H.G.
        • Luigjes J.
        • van den Heuvel M.P.
        • Cahn W.
        • Kahn R.S.
        • Hulshoff Pol H.E.
        Tract-based analysis of magnetization transfer ratio and diffusion tensor imaging of the frontal and frontotemporal connections in schizophrenia.
        Schizophr Bull. 2010; 36: 778-787
        • Tsai G.
        • Coyle J.T.
        N-acetylaspartate in neuropsychiatric disorders.
        Prog Neurobiol. 1995; 46: 531-540
        • Lim K.O.
        • Adalsteinsson E.
        • Spielman D.
        • Sullivan E.V.
        • Rosenbloom M.J.
        • Pfefferbaum A.
        Proton magnetic resonance spectroscopic imaging of cortical gray and white matter in schizophrenia.
        Arch Gen Psychiatry. 1998; 55: 346-352
        • Ongur D.
        • Prescot A.P.
        • Jensen J.E.
        • Rouse E.D.
        • Cohen B.M.
        • Renshaw P.F.
        • et al.
        T2 relaxation time abnormalities in bipolar disorder and schizophrenia.
        Magn Reson Med. 2010; 63: 1-8
        • Ackerman J.J.
        • Neil J.J.
        The use of MR-detectable reporter molecules and ions to evaluate diffusion in normal and ischemic brain.
        NMR Biomed. 2010; 23: 725-733
        • Upadhyay J.
        • Hallock K.
        • Ducros M.
        • Kim D.S.
        • Ronen I.
        Diffusion tensor spectroscopy and imaging of the arcuate fasciculus.
        Neuroimage. 2008; 39: 1-9
        • Harada M.
        • Uno M.
        • Hong F.
        • Hisaoka S.
        • Nishitani H.
        • Matsuda T.
        Diffusion-weighted in vivo localized proton MR spectroscopy of human cerebral ischemia and tumor.
        NMR Biomed. 2002; 15: 69-74
        • Abe O.
        • Okubo T.
        • Hayashi N.
        • Saito N.
        • Iriguchi N.
        • Shirouzu I.
        • et al.
        Temporal changes of the apparent diffusion coefficients of water and metabolites in rats with hemispheric infarction: Experimental study of transhemispheric diaschisis in the contralateral hemisphere at 7 tesla.
        J Cereb Blood Flow Metab. 2000; 20: 726-735
        • Dreher W.
        • Busch E.
        • Leibfritz D.
        Changes in apparent diffusion coefficients of metabolites in rat brain after middle cerebral artery occlusion measured by proton magnetic resonance spectroscopy.
        Magn Reson Med. 2001; 45: 383-389
        • Liu Z.
        • Liu X.
        • Hui L.
        • Zhao D.
        • Wang X.
        • Xie S.
        • et al.
        The appearance of ADCs in the non-affected areas of the patients with MELAS.
        Neuroradiology. 2011; 53: 227-232
        • Chomiak T.
        • Hu B.
        What is the optimal value of the g-ratio for myelinated fibers in the rat CNS? A theoretical approach.
        PLoS One. 2009; 4: e7754
        • Kandel E.R.
        • Schwartz J.H.
        • Jessell T.M.
        Principles of Neural Science.
        3rd ed. Elsevier, New York1991
        • Paus T.
        • Toro R.
        Could sex differences in white matter be explained by g ratio?.
        Front Neuroanat. 2009; 3: 14
        • Nave K.A.
        Myelination and the trophic support of long axons.
        Nat Rev Neurosci. 2010; 11: 275-283
        • Uranova N.A.
        • Vostrikov V.M.
        • Vikhreva O.V.
        • Zimina I.S.
        • Kolomeets N.S.
        • Orlovskaya D.D.
        The role of oligodendrocyte pathology in schizophrenia.
        Int J Neuropsychopharmacol. 2007; 10: 537-545
        • Flynn S.W.
        • Lang D.J.
        • Mackay A.L.
        • Goghari V.
        • Vavasour I.M.
        • Whittall K.P.
        • et al.
        Abnormalities of myelination in schizophrenia detected in vivo with MRI, and post-mortem with analysis of oligodendrocyte proteins.
        Mol Psychiatry. 2003; 8: 811-820
        • McCullumsmith R.E.
        • Gupta D.
        • Beneyto M.
        • Kreger E.
        • Haroutunian V.
        • Davis K.L.
        • et al.
        Expression of transcripts for myelination-related genes in the anterior cingulate cortex in schizophrenia.
        Schizophr Res. 2007; 90: 15-27
        • Hakak Y.
        • Walker J.R.
        • Li C.
        • Wong W.H.
        • Davis K.L.
        • Buxbaum J.D.
        • et al.
        Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia.
        Proc Natl Acad Sci U S A. 2001; 98: 4746-4751
        • Budel S.
        • Padukkavidana T.
        • Liu B.P.
        • Feng Z.
        • Hu F.
        • Johnson S.
        • et al.
        Genetic variants of Nogo-66 receptor with possible association to schizophrenia block myelin inhibition of axon growth.
        J Neurosci. 2008; 28: 13161-13172
        • Bernstein H.G.
        • Steiner J.
        • Bogerts B.
        Glial cells in schizophrenia: Pathophysiological significance and possible consequences for therapy.
        Expert Rev Neurother. 2009; 9: 1059-1071
        • de Graaf R.A.
        • Luo Y.
        • Garwood M.
        • Nicolay K.
        B1-insensitive, single-shot localization and water suppression.
        J Magn Reson B. 1996; 113: 35-45
      1. Chen W, Luo Y, Merkle H, Zhu X-H, Adriany G, Garwood M, et al. (1995): Low-power, B1-insensitive, frequency-selective saturation pulse for use of water suppression and saturation transfer in NMR. SMR, 3rd Annual Meeting. Nice, France: Society for Magnetic Resonance, 1016.

        • Kan H.E.
        • Techawiboonwong A.
        • van Osch M.J.
        • Versluis M.J.
        • Deelchand D.K.
        • Henry P.G.
        • et al.
        Differences in apparent diffusion coefficients of brain metabolites between grey and white matter in the human brain measured at 7 T.
        Magn Reson Med. 2012; 67: 1203-1209
        • Tkac I.
        • Starcuk Z.
        • Choi I.Y.
        • Gruetter R.
        In vivo 1H NMR spectroscopy of rat brain at 1 ms echo time.
        Magn Reson Med. 1999; 41: 649-656
        • Choi J.K.
        • Dedeoglu A.
        • Jenkins B.G.
        Application of MRS to mouse models of neurodegenerative illness.
        NMR Biomed. 2007; 20: 216-237
        • Camchong J.
        • Lim K.O.
        • Sponheim S.R.
        • Macdonald A.W.
        Frontal white matter integrity as an endophenotype for schizophrenia: Diffusion tensor imaging in monozygotic twins and patients’ nonpsychotic relatives.
        Front Hum Neurosci. 2009; 3: 35
        • Valette J.
        • Guillermier M.
        • Besret L.
        • Hantraye P.
        • Bloch G.
        • Lebon V.
        Isoflurane strongly affects the diffusion of intracellular metabolites, as shown by 1H nuclear magnetic resonance spectroscopy of the monkey brain.
        J Cereb Blood Flow Metab. 2007; 27: 588-596
        • Chakraborty G.
        • Mekala P.
        • Yahya D.
        • Wu G.
        • Ledeen R.W.
        Intraneuronal N-acetylaspartate supplies acetyl groups for myelin lipid synthesis: Evidence for myelin-associated aspartoacylase.
        J Neurochem. 2001; 78: 736-745
        • Ben-Shachar D.
        • Laifenfeld D.
        Mitochondria, synaptic plasticity, and schizophrenia.
        Int Rev Neurobiol. 2004; 59: 273-296
        • Du F.
        • Cooper A.
        • Cohen B.M.
        • Renshaw P.F.
        • Ongur D.
        Water and metabolite transverse T2 relaxation time abnormalities in the white matter in schizophrenia.
        Schizophr Res. 2012; 137: 241-245
        • Maudsley A.A.
        • Domenig C.
        • Govind V.
        • Darkazanli A.
        • Studholme C.
        • Arheart K.
        • et al.
        Mapping of brain metabolite distributions by volumetric proton MR spectroscopic imaging (MRSI).
        Magn Reson Med. 2009; 61: 548-559
        • Posse S.
        • Otazo R.
        • Caprihan A.
        • Bustillo J.
        • Chen H.
        • Henry P.G.
        • et al.
        Proton echo-planar spectroscopic imaging of J-coupled resonances in human brain at 3 and 4 Tesla.
        Magn Reson Med. 2007; 58: 236-244
        • Coyle J.T.
        The nagging question of the function of N-acetylaspartylglutamate.
        Neurobiol Dis. 1997; 4: 231-238
        • Pouwels P.J.
        • Frahm J.
        Differential distribution of NAA and NAAG in human brain as determined by quantitative localized proton MRS.
        NMR Biomed. 1997; 10: 73-78
        • Govindaraju V.
        • Young K.
        • Maudsley A.A.
        Proton NMR chemical shifts and coupling constants for brain metabolites.
        NMR in Biomedicine. 2000; 13: 129-153
        • Laule C.
        • Vavasour I.M.
        • Kolind S.H.
        • Li D.K.
        • Traboulsee T.L.
        • Moore G.R.
        • et al.
        Magnetic resonance imaging of myelin.
        Neurotherapeutics. 2007; 4: 460-484
        • MacKay A
        • Laule C
        • Vavasour I
        • Bjarnason T
        • Kolind S
        • Madler B
        Insights into brain microstructure from the T2 distribution.
        Magn Reson Imaging. 2006; 24: 515-525
        • Friedman J.I.
        • Tang C.
        • Carpenter D.
        • Buchsbaum M.
        • Schmeidler J.
        • Flanagan L.
        • et al.
        Diffusion tensor imaging findings in first-episode and chronic schizophrenia patients.
        Am J Psychiatry. 2008; 165: 1024-1032
      2. Heng S, Song AW, Sim K (2010): White matter abnormalities in bipolar disorder: Insights from diffusion tensor imaging studies. J Neural Transm 117:639–654.

        • White T.
        • Nelson M.
        • Lim K.O.
        Diffusion tensor imaging in psychiatric disorders.
        Top Magn Reson Imaging. 2008; 19: 97-109

      Linked Article

      • Diving Deep into White Matter to Improve Our Understanding of the Pathophysiology of Schizophrenia
        Biological PsychiatryVol. 74Issue 6
        • Preview
          Innovations in our capacity to image the fine structure of the human brain are urgently needed because after more than a century of research into the biology of schizophrenia, we still find ourselves far from understanding the neurobiological mechanisms responsible for the disorder. Without such an understanding, we are unlikely to discover more effective ways of preventing or reversing the underlying pathology. In this issue of Biological Psychiatry, Du et al. (1) describe the use of two in magnetic resonance (MR)-based technologies that, when used in combination, allowed them to make specific references about the nature of white matter abnormalities in patients with schizophrenia.
        • Full-Text
        • PDF