- Email: [email protected]
Correlation Between Serum Brain-Derived Neurotrophic Factor Level and An In Vivo Marker of Cortical Integrity
Correlation Between Serum Brain-Derived Neurotrophic Factor Level and An In Vivo Marker of Cortical Integrity Undine E. Lang, Rainer Hellweg, Frank Seifert, Florian Schubert, and Juergen Gallinat Background: Brain-derived neurotrophic factor (BDNF) signaling at synapses improves synaptic strengthening associated with learning and memory. In the present study we hypothesized that serum BDNF concentration is associated with in vivo level of cerebral Nacetylaspartate (NAA), a well established marker of neuronal integrity. Methods: In 36 healthy subjects BDNF serum concentration and absolute concentration of NAA together with other metabolites were measured by proton magnetic resonance spectroscopy (1H-MRS) in regions with high BDNF levels (anterior cingulate cortex [ACC], left hippocampus). Relationship between BDNF concentration and brain metabolites was studied in linear regression analysis with BDNF concentration as dependent variable and metabolite concentrations, age, and gender as predictor variables. Results: The BDNF serum concentrations were positively associated with the concentrations of NAA (T ⫽ 2.193, p ⫽ .037) and total choline (T ⫽ 1.997, p ⫽ .055; trend) but not total creatine or glutamate in the ACC. No significant association was observed between BDNF serum concentration and absolute metabolite concentrations in the hippocampus. Conclusions: The preliminary data might indicate that BDNF serum concentration reflects some aspects of neuronal plasticity as indicated by its association with NAA level in the cerebral cortex. The results would be in line with the notion that BDNF plays a central role in the regulation of neuronal survival and differentiation in the human brain. Key Words: BDNF, brain-derived neurotrophic factor, cortical integrity, glutamate, NAA, proton magnetic resonance spectroscopy (1H-MRS)
B
rain-derived neurotrophic factor (BDNF) signaling at synapses enhances neuronal plasticity and long-term potentiation (LTP), which represents a process of synaptic strengthening associated with learning and memory (Ernfors and Braham 2003; Figurov et al. 1996; Poo, 2001; Rex et al. 2006). Likewise, long-term developmental phenomena, such as neuronal survival, migration, and differentiation are mediated by BDNF through its ability to promote activity-dependent refinement of synaptic architecture (Baquet et al. 2004; Gorski et al. 2003; Poo 2001). The highest levels of BDNF are observed in hippocampus and cerebral cortex (Phillips et al. 1990), brain regions that are predominantly involved in higher cognitive functions (e.g., Gallinat et al. 2006). Expression of BDNF in these brain regions has been implicated in neurodevelopmental and neurodegenerative disorders, including Alzheimer’s disease, Huntington’s disease, and Down’s syndrome (Baquet et al. 2004; Bimonte-Nelson et al. 2003; Weickert et al. 2003; Zuccato et al. 2005), and abnormal BDNF concentrations in the anterior cingulate cortex (ACC) and hippocampus have been postulated as molecular substrate for pathological manifestations of schizophrenia (Takahashi et al. 2000). Because it has been described that intact BDNF in the peripheral circulation crosses the blood-brain barrier (Pan et al.
From the Department of Psychiatry and Psychotherapy (UEL, JG), Charité – University Medicine Berlin, Campus Mitte; Department of Psychiatry (RH), Charité—University Medicine Berlin, Campus Benjamin Franklin; and Physikalisch-Technische Bundesanstalt (FS, FS), Berlin, Germany. Address reprint requests to J. Gallinat, M.D., Ph.D., Charité Medicine Berlin, Clinic for Psychiatry and Psychotherapy, St. Hedwig Krankenhaus, Turmstrasse 21, 10559 Berlin, Germany; E-mail: [email protected]. Received October 12, 2006; revised January 3, 2007; accepted January 3, 2007.
0006-3223/07/$32.00 doi:10.1016/j.biopsych.2007.01.002
1998), BDNF blood concentrations might reflect central BDNF concentrations. However, an association of serum BDNF concentration with in vivo measured parameters of neuronal plasticity in humans has not been investigated so far. Although such a correlation would not prove a causal relationship between parameters, it would represent an argument for a cerebral role of peripheral measured BDNF. N-acetylaspartate (NAA) is a well established biochemical marker of neuronal damage, and its measurement in the acute phase of brain injury indicates the degree of neuronal loss and therefore reflects tissue metabolic dysfunction (Davie et al. 1997; Ebisu et al. 1994; Sager et al. 2001). Because cellular dysfunction can cause stronger reductions in NAA level than neuronal loss, NAA quantification has been suggested as a potential tool for visualizing the penumbra area in stroke patients (Demougeot et al. 2004). The NAA changes in mild cognitive impairment, Alzheimer’s disease (Ackl et al. 2005), and Parkinson=s disease (Lucetti et al. 2001) have been described. Apart from pathological conditions, cerebral NAA concentrations measured in healthy subjects have been associated with memory tasks (Gimenez et al. 2004) and intellectual performance and intelligence (Jung et al. 2005) as well as personality traits (Gallinat et al. 2005). Moreover, a use-dependent increase of cortical NAA has been described suggesting that this parameter might reflect neuronal adaption and plasticity in humans (Aydin et al. 2005). The technique of proton magnetic resonance spectroscopy (1H-MRS) permits the quantification of a number of brain metabolites noninvasively; with regard to sensitivity and spectral resolution (i.e., metabolite selectivity), it has benefited to a large extent from the increasing field strength of magnetic resonance (MR) scanners (e.g., Bartha et al. 2000; Kim and Garwood 2003; Tkac et al. 2001). Thus, using a field strength ⬎ 1.5-Tesla not only improves the signal-to-noise ratio for the commonly detected singlet resonances of NAA, choline-containing compounds (tCho), and creatine ⫹ phosphocreatine (tCr) but also permits, in principle, the estimation of the levels of other metabolites, such BIOL PSYCHIATRY 2007;62:530 –535 © 2007 Society of Biological Psychiatry
U.E. Lang et al. as neurotransmitters (Theberge et al. 2002), Vitamin C (Terpstra and Gruetter 2004), or reduced glutathione (Trabesinger et al. 1999). Because we have developed a methodology for the quantification of the aforementioned singlets together with Lglutamate at 3-Tesla (Gallinat et al. 2006; Schubert et al. 2004), we used this approach to test the hypothesis that BDNF serum concentration is associated with in vivo measured brain NAA as a marker of neuronal integrity in healthy subjects. In an exploratory manner the relationships of BDNF to other metabolic parameters (tCho, tCr, glutamate) were investigated. The present measurements were performed in regions displaying high tissue concentrations of BDNF, namely the ACC and the left hippocampus (Schulte-Herbruggen et al. 2006). It was hypothesized that NAA would show a positive correlation with BDNF serum concentration, because BDNF has been implicated in the structure and function of the adult brain (Bath and Lee 2006).
Methods and Materials Subjects The study was approved by the ethics committee of the Charité University Medicine Berlin. Thirty-six healthy subjects (age: 34.4 ⫾ 10.1 years, 17 men 33.8 ⫾ 8.9 years, 19 women 34.9 ⫾ 11.2 years) were recruited through newspaper advertisements. They were of German descent and gave written informed consent. All subjects were free of medical, neurological, and psychiatric disorders as determined by in-person interviews (Mini-International Neuropsychiatric Interview [M.I.N.I.]; Sheehan et al. 1998) performed by a psychiatrist. Subjects with a family history (first degree) of axis I disorder were excluded from participation. For further methodological details, see Gallinat et
BIOL PSYCHIATRY 2007;62:530 –535 531 al. (2002) and Lang et al. (2005). Blood for BDNF determination in serum was collected before the 1H-MRS investigation. MR Spectroscopy Magnetic resonance measurements were carried out on a 3-Tesla scanner (MEDSPEC 30/100; Bruker Biospin, Ettlingen, Germany) with a circularly polarized head coil. After automated global shim of the linear, xz, z2, and x2-y2 field components, T1-weighted images (modified driven equilibrium Fourier transform [MDEFT], echo time (TE) ⫽ 5.5 msec, repetition time (TR) ⫽ 23.4 msec, 64 contiguous slices of 2-mm thickness, 1-mm inplane (x-y) resolution) were acquired. The MR spectra were recorded from 2 ⫻ 3 ⫻ 2 cm3 voxels including the left hippocampus and from 2.5 ⫻ 4 ⫻ 2 cm3 voxels including the ACC of the volunteer brains (Figure 1). For metabolite quantification we closely followed an established method (Schubert et al. 2004). After manual shimming to water linewidths (full width at half maximum [FWHM]) of 7–9 Hz and 6 –7 Hz for the hippocampus and ACC voxels, respectively, and determination of the radiofrequency power needed for a 90° excitation pulse, calibration of water suppression was carried out, followed by acquisition of spectra with point resolved spectroscopy (PRESS). To obtain one MR spectrum, eight subspectra of 16 phase cycled scans each were recorded with TR ⫽ 3 sec and TE ⫽ 80 msec, giving 128 averages. Subsequently the water suppression pulses were switched off to acquire a waterunsuppressed spectrum (n ⫽ 8). Before further processing, the eight individual metabolite subspectra were corrected for eddy currents with the water-unsuppressed spectrum and automatically frequency aligned to correct for frequency shifts during the
Figure 1. Voxel positions shown on typical brain modified driven equilibrium Fourier transform (MDEFT) images; anterior cingulate cortex (left), hippocampus (right).
www.sobp.org/journal
532 BIOL PSYCHIATRY 2007;62:530 –535
Figure 2. Point resolved spectroscopy (PRESS) spectra (repetition time [TR] ⫽ 3 sec, and echo time [TE] ⫽ 80 msec) acquired from voxels containing the anterior cingulate cortex (ACC) (top) and left hippocampus (bottom). tCr, total creatine; tCho, total choline; NAA, N-acetylaspartate; Glu, glutamate. Note that, apart from phasing, no postprocessing was carried out.
scan time (of more than 6 min), caused by involuntary subject motion and system instabilities. Examples of resulting spectra— then of 128 averages—are shown in Figure 2. The spectral linewidth (FWHM) values of the metabolite resonances are estimated by the used fitting procedure (described in following text); mean FWHM (SD) (in Hz) were 4.9 (.6) for the ACC voxel and 6.6 (.9) for the hippocampus voxel and thus below the minimum acceptable linewidth for high-precision MR spectroscopy suggested recently (Kreis 2004). Spectra were quantified with a program package that relies on a time domain-frequency domain fitting procedure and involves inclusion of phantom basis spectra and prior knowledge into the fit (Elster et al. 2005; Schubert et al. 2004). At the chosen TE of 80 msec the baseline was not entirely flat. To account for baseline features, the fitting procedure routinely includes estimation of the baseline nonparametrically by regularization. In the present spectra, total choline (tCho), total creatine (tCr), NAA, glutamate (Glu), and glutamine resonances were fitted by inclusion of phantom spectra for the latter three and imposing constant frequency differences for glutamate, glutamine, and NAA and equal linewidths as prior knowledge. Extensive tests yielded mean uncertainties (corresponding to Cramér-Rao lower bounds with added uncertainties from background modeling) for the fitting of NAA of 2.4% and 2.7%, tCho 3.7% and 4.0%, tCr 3.0% and 3.1%, and Glu 11.0% and 13.2% for the ACC and the hippocampus voxel, respectively. Metabolite amplitudes returned by the fitting procedure were corrected for different coil loading by the phantoms and the individual subject’s head (principle of reciprocity [Danielsen and Henriksen 1994]) and for relaxation effects, assuming no differences in relaxation behavior between the www.sobp.org/journal
U.E. Lang et al. subjects. Transverse relaxation times were determined from three healthy volunteers with TE of 50, 80, 135, 250, and 330 msec. The results (in msec, SD in parentheses) for Glu(C4), NAA, tCr, and tCho were 194 (37), 278 (31), 179 (9), and 282 (45); and 171 (22), 267 (15), 198 (31), and 291 (13) for the ACC voxel and for the hippocampus voxel, respectively. It should be noted that, although perfectly applicable for the correction, the transverse relaxation time of Glu is only a relative one, because it was determined by fitting of the in vivo data to phantom spectra acquired at the same TE. As T1 for NAA, tCr, and tCho, the following mean values determined from two healthy volunteers with TE ⫽ 30 msec and TR ⫽ .5, 1, 2, 3, 5, and 7 s were employed in the analysis: 1.48 sec; 1.21 sec; and 1.44 sec. For Glu, no correction for longitudinal relaxation effects was carried out. The deviation caused by the T1 effect at TR ⫽ 3 sec, and an assumed T1 of glutamate at 3-T of approximately 1.2 sec (Mlynarik et al. 2001) was largely compensated by the deviation estimated for the aqueous, buffered glutamate phantom, where TR was set at 5 sec and T1 was determined to be 1.47 sec. Thus, glutamate might at worst be systematically underestimated by about 5%, owing to unaccounted T1 effects, which would lead to uniform scaling of the calculated concentrations. Although necessary for a stable Glu quantification, the inclusion of glutamine in the fit gave only an estimate of its concentration, the uncertainty of which was too large for further consideration (Schubert et al. 2004). The in vivo concentrations were corrected for the content of cerebrospinal fluid (CSF) of the voxels studied, as derived from segmentation of the T1-weighted images with spm99 (Ashburner and Friston 1997). Providing the highest retest reliability (Schubert et al. 2004), a pixel was classified depending on which spm99 tissue classification had the greatest probability. Serum aliquots for neurotrophin measurements were stored at ⫺70° C before batch processing. Levels of endogenous BDNF were measured in diluted serum samples by a highly sensitive and specific fluorometric two-site enzyme-linked immunosorbent assay with commercial kits, in principle according to the manufacturer’s instructions (Promega, Mannheim, Germany) but in an improved, fluorometric form described previously (Hellweg et al. 2003, Ziegenhorn et al. 2006). Statistical Analyses To study the relationships between BDNF serum concentrations and brain metabolites, linear regression analysis was performed with the independent factors brain metabolites (NAA, tCho, tCr, Glu), age, and gender and the dependent variable BDNF serum concentration. Linear regression analyses were separately computed for hippocampal and ACC metabolite concentrations. For correlation analyses, Pearson correlation coefficients were computed. All tests were performed at a level of significance of p ⫽ .05.
Results The mean BDNF serum concentration amounted to 15.8 ⫾ 8.4 ng/mL. The absolute metabolite concentrations (in mmol/L) in the ACC were 13.6 ⫾ 1.2 for NAA, 2.14 ⫾ .25 for total choline (tCho), 9.4 ⫾ 1.0 for total creatine (tCr), and 11.5 ⫾ 1.6 for Glu. In the hippocampus the concentrations (in mmol/L) were 11.6 ⫾ 1.0 for NAA, 2.21 ⫾ .26 for tCho, 9.6 ⫾ .6 for tCr, and 10.9 ⫾ 1.2 for Glu. The linear regression analysis revealed that BDNF serum concentration was predicted by the concentrations of NAA (T ⫽ 2.193, p ⫽ .037) and, as a statistical trend, tCho (T ⫽ 1.997, p ⫽ .055) measured in the ACC. The result with regard to NAA
U.E. Lang et al.
BIOL PSYCHIATRY 2007;62:530 –535 533
Figure 3. Scatterplots of NAA concentrations measured in the ACC versus serum brain-derived neurotrophic factor (BDNF) concentrations (A) and their residuals after linear correction for age and gender (B). Significant positive correlations were observed in both analyses (r ⫽ .396, p ⫽ .017 and r ⫽ .448, p ⫽ .008, respectively). Other abbreviations as in Figure 2.
remained significant (T ⫽ 2.111, p ⫽ .044) when the regression analysis was repeated with the additional factor gray matter content of the voxel (T ⫽ ⫺.190, p ⫽ .851). No other metabolite concentration (i.e., tCr or Glu) was a significant predictor, nor were age or gender. The linear regression analysis of BDNF concentration with the hippocampal metabolite concentrations as predictor variables did not reveal significant effects. This was also the case when the gray matter content of the voxel was included in the analysis. A correlation analysis was carried out for the brain metabolite significantly predicting the BDNF concentration. Pearson correlation analysis showed a significant positive correlation between BDNF concentration and the concentration of NAA in the ACC (r ⫽ .396, p ⫽ .017; Figure 3A). This association was also significant when the correlation analysis was controlled for age and gender (r ⫽ .448, p ⫽ .008, partial correlation; Figure 3B).
Discussion We found a positive correlation between cortical NAA concentration and BDNF serum concentration in healthy human subjects. No significant relationships between BDNF serum concentrations and other cortical or any hippocampal brain metabolites were observed. The results indicate that neuronal integrity and vitality of a cortical region like the ACC, as reflected by a high concentration of NAA, might be reflected by high serum concentrations of BDNF. Brain-derived neurotrophic factor (BDNF), like other neurotrophins, has long-term effects on neuronal survival and differentiation (Poo 2001). Thus prefrontally targeted BDNF knock-out mice have been shown to display reductions in cortical volume during adulthood (Baquet et al. 2004); an age-related volume reduction was found in the dorsolateral prefrontal cortices in humans carrying the Met-allele of the BDNF gene (Nemoto et al. 2006). These findings indicate a role of BDNF for the morphology and integrity of cortical brain areas that is in line with the positive correlation between serum BDNF and cortical NAA concentrations observed in our study. Further evidence indicates that the cortical level of NAA also reflects brain function (Bertolino et al. 1999) as well as behavioral aspects or personality traits
(see opening text of this report). For instance, in an in vivo 1H-MRS study an association was found between cortical NAA (ACC) and the level of cognitive interference in the Stroop Color-Word task (Grachev et al. 2001), a paradigm known to be dependent on ACC function (Bench et al. 1993). Grachev et al. suggested their results to be a correlate of a dynamic process of neuronal reorganization in the ACC resulting from use-dependent activity. This view is compatible with a recently reported use-dependent cortical NAA increase in humans (Aydin et al. 2005) and with findings showing neurotrophins to rapidly increase spontaneous firing rate and synaptic transmission in cortical neurons (He et al. 2005). The BDNF signaling at synapses enhances activity-dependent long-term potentiation (Ernfors and Bramham 2003; Figurov et al. 1996; Levine and Kolb 2000; Poo 2001), a process relevant for long-term plasticity in the ACC. In this context, the present results might indicate that high cortical NAA concentrations reflect ongoing neuronal plasticity mediated by BDNF. Interestingly, the majority of reports describe an association of BDNF with functional (Egan et al. 2003; Hariri et al. 2003) and structural aspects (Bueller et al. 2006) of the hippocampus, a brain region with a high expression of BDNF protein. In the present study, no association between BDNF serum concentration and hippocampal neurochemistry was observed, even when the analysis was controlled for the gray matter content of the hippocampal voxel. There could be several reasons for this result: for instance, the relatively large voxel size, including regions other than the hippocampus, and a slightly higher uncertainty of metabolite quantification compared with that in the ACC (Schubert et al. 2004). Another confounding factor might be the reported effect of a missense variation of the BDNF gene (Val66Met substitution) on the level of NAA in the hippocampus of healthy volunteers (Egan et al. 2003). However, the most important reason for a lacking correlation between BDNF and hippocampal metabolites might be the relatively small size of the hippocampus compared with the cerebral cortex in relation to whole brain volume. Given that cerebral BDNF crosses the blood-brain barrier (Pan et al. 1998), it is reasonable to assume that serum BDNF concentrations are associated more closely www.sobp.org/journal
534 BIOL PSYCHIATRY 2007;62:530 –535 with neocortical than with hippocampal BDNF levels. This assumption would be in line with the finding in an animal experiment that BDNF in serum is correlated with BDNF expression in cortical but not hippocampal region (Karege et al. 2002). Other evidence indicates that BDNF as well as NAA concentrations are modulated by pharmacological intervention more dynamically in the ACC (or cortical regions) than in the hippocampus (Rantamaki et al. 2006; Yildiz-Yesiloglu and Ankerst 2006). Dynamic BDNF regulation might play a role for the observed positive correlation between cortical NAA and serum BDNF. In a BDNF mutant mouse model the hypothesis has been worked out that BDNF supports the maintenance of cortical neuron size and dendrite structure rather than the initial development of these features (Gorski et al. 2003). For this reason we assume permanent BDNF support to be required at least in cortical areas to maintain neuronal functioning and suggest NAA to be a marker of this process. This model is also in line with an age-dependent decline of serum and cortical BDNF levels (Webster et al. 2006; Ziegenhorn et al. 2006) as well as cerebral NAA (Schubert et al. 2004). Of note, BDNF concentration in human hippocampus does not change significantly over the life span (Webster et al. 2006). However, other arguments contradict the role of NAA as a marker of neuronal plasticity. For instance, the concentration of NAA in infants during development is lower compared with adults (Kadota et al. 2001). Furthermore, several biological conditions, like menstrual cycle or body weight, have an effect on BDNF concentrations in the peripheral blood (Lommatzsch et al. 2005), whereas cerebral levels of NAA are largely unaffected by these factors (Epperson et al. 2005). However, the effect sizes of the mentioned biological factors on BDNF concentration are rather small, which might not completely exclude a relationship between BDNF and the cerebral metabolite NAA. In principle, the use of the current datasets in previously published papers (Gallinat et al. 2005, 2006, 2007; Schubert et al. 2004) increases the risk of false positive results, which have to be taken into account when interpreting the actual results. A trend was observed for the concentrations of cortical tCho to increase with increased serum BDNF. Although not significant, this is compatible with a dynamic maintenance of neurons driven by cortical BDNF, because tCho fluctuates when cellular membranes are degraded or rapidly synthesized (Dawson 1985; Porcellati and Arienti 1983). We did not find a significant correlation between BDNF serum concentration and cerebral glutamate, although both components have previously been shown to interact in several ways (Bustos et al. 2004). However, the precision of cerebral glutamate measurement is lower than that for NAA (Schubert et al. 2004), which might hinder the detection of a relationship with serum BDNF. In conclusion, we detected for the first time an association between serum BDNF protein concentration and cortical NAA concentration in the living human brain. The results are in line with the formerly described role of neurotrophins as regulatory factors mediating the differentiation and survival of neurons (Poo 2001) as reflected by the level of NAA.
Support from the German Federal Ministry of Education and Research (BMBF, Project Berlin Neuroimaging Center, No. 01G00208) is gratefully acknowledged. Drs. Lang and Hellweg contributed equally to this paper. www.sobp.org/journal
U.E. Lang et al. Ackl N, Ising M, Schreiber YA, Atiya M, Sonntag A, Auer DP (2005): Hippocampal metabolic abnormalities in mild cognitive impairment and Alzheimer’s disease. Neurosci Lett 384:23–28. Ashburner J, Friston K (1997): Multimodal image coregistration and partitioning—a unified framework. Neuroimage 6:209 –217. Aydin K, Ciftci K, Terzibasioglu E, Ozkan M, Demirtas A, Sencer S, et al.(2005): Quantitative proton MR spectroscopic findings of cortical reorganization in the auditory cortex of musicians. AJNR Am J Neuroradiol 26:128 – 136. Baquet ZC, Gorski JA, Jones KR (2004): Early striatal dendrite deficits followed by neuron loss with advanced age in the absence of anterograde cortical brain-derived neurotrophic factor. J Neurosci 24:4250 – 4258. Bartha R, Drost DJ, Menon RS, Williamson PC (2000): Comparison of the quantification precision of human short echo time (1)H spectroscopy at 1.5 and 4.0 Tesla. Magn Reson Med 44:185–192. Bath KG, Lee FS (2006): Variant BDNF (Val66Met) impact on brain structure and function. Cogn Affect Behav Neurosci 6:79 – 85. Bench CJ, Frith CD, Grasby PM, Friston KJ, Paulesu E, Frackowiak RS, et al.(1993): Investigations of the functional anatomy of attention using the Stroop test. Neuropsychologia 31:907–922. Bertolino A, Knable MB, Saunders RC, Callicott JH, Kolachana B, Mattay VS, et al.(1999): The relationship between dorsolateral prefrontal N-acetylaspartate measures and striatal dopamine activity in schizophrenia. Biol Psychiatry 45:660 – 667. Bimonte-Nelson HA, Hunter CL, Nelson ME, Granholm AC (2003): Frontal cortex BDNF levels correlate with working memory in an animal model of Down syndrome. Behav Brain Res 139:47–57. Bueller JA, Aftab M, Sen S, Gomez-Hassan D, Burmeister M, Zubieta JK (2006): BDNF Val(66)Met al.ele is associated with reduced hippocampal volume in healthy subjects. Biol Psychiatry 59:812– 815. Bustos G, Abarca J, Campusano J, Bustos V, Noriega V, Aliaga E (2004): Functional interactions between somatodendritic dopamine release, glutamate receptors and brain-derived neurotrophic factor expression in mesencephalic structures of the brain. Brain Res Brain Res Rev 47:126 – 144. Danielsen ER, Henriksen O (1994): Absolute quantitative proton NMR spectroscopy based on the amplitude of the local water suppression pulse. Quantification of brain water and metabolites. NMR Biomed 7:311–318. Davie CA, Barker GJ, Thompson AJ, Tofts PS, McDonald WI, Miller DH (1997): 1H magnetic resonance spectroscopy of chronic cerebral white matter lesions and normal appearing white matter in multiple sclerosis. J Neurol Neurosurg Psychiatry 63:736 –742. Dawson RMC (1985): Enzymatic pathways of phospholipid metabolism in the nervous system. In: Eichberg J, editor. Phospholipids in Nervous Tissues. New York: Wiley, 45–78. Demougeot C, Marie C, Giroud M, Beley A (2004): N-acetylaspartate: A literature review of animal research on brain ischaemia. J Neurochem 90: 776 –783. Ebisu T, Rooney WD, Graham SH, Weiner MW, Maudsley AA (1994): N-acetylaspartate as an in vivo marker of neuronal viability in kainate-induced status epilepticus: 1H magnetic resonance spectroscopic imaging. J Cereb Blood Flow Metab 14:373–382. Egan MF, Kojima M, Callicott JH, Goldberg TE, Kolachana BS, Bertolino A, et al.(2003): The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 112:257–269. Elster C, Schubert F, Link A, Walzel M, Seifert F, Rinneberg H (2005): Quantitative magnetic resonance spectroscopy: Semi-parametric modeling and determination of uncertainties. Magn Reson Med 53:1288 –1296. Epperson CN, O’Malley S, Czarkowski KA, Gueorguieva R, Jatlow P, Sanacora G, et al.(2005): Sex, GABA, and nicotine: The impact of smoking on cortical GABA levels across the menstrual cycle as measured with proton magnetic resonance spectroscopy. Biol Psychiatry 57:44 – 48. Ernfors P, Bramham CR (2003): The coupling of a trkB tyrosine residue to LTP. Trends Neurosci 26:171–173. Figurov A, Pozzo-Miller LD, Olafsson P, Wang T, Lu B (1996): Regulation of synaptic responses to high-frequency stimulation and LTP by neurotrophins in the hippocampus. Nature 381:706 –709. Gallinat J, Kunz D, Lang UE, Neu P, Kassim N, Kienast T, et al.(2007): Association between cerebral glutamate and human behaviour: The sensation seeking personality trait. Neuroimage 34:671– 678.
U.E. Lang et al. Gallinat J, Kunz D, Senkowski D, Kienast T, Seifert F, Schubert F, et al.(2006): Hippocampal glutamate concentration predicts cerebral theta oscillations in human memory utilization. Psychopharmacology 187:103–111. Gallinat J, Mulert C, Bajbouj M, Herrmann WM, Schunter J, Senkowski D, et al.(2002): Frontal and temporal dysfunction of auditory stimulus processing in schizophrenia. Neuroimage 17:110 –127. Gallinat J, Ströhle A, Lang UE, Bajbouj M, Kalus P, Montag C, et al.(2005): Association of human hippocampal neurochemistry, serotonin transporter genetic variation, and anxiety. Neuroimage 26:123–131. Gimenez M, Junque C, Narberhaus A, Caldu X, Segarra D, Vendrell P, et al.(2004): Medial temporal MR spectroscopy is related to memory performance in normal adolescent subjects. Neuroreport 15:703–707. Gorski JA, Zeiler SR, Tamowski S, Jones KR (2003): Brain-derived neurotrophic factor is required for the maintenance of cortical dendrites. J Neurosci 23:6856 – 6865. Grachev ID, Kumar R, Ramachandran TS, Szeverenyi NM (2001): Cognitive interference is associated with neuronal marker N-acetyl aspartate in the anterior cingulate cortex: An in vivo (1)H-MRS study of the Stroop ColorWord task. Mol Psychiatry 6:529 –539. Hariri AR, Goldberg TE, Mattay VS, Kolachana BS, Callicott JH, Egan MF, et al.(2003): Brain-derived neurotrophic factor val66met polymorphism affects human memory related hippocampal activity and predicts memory performance. J Neurosci 23:6690 – 6694. He J, Gong H, Luo Q (2005): BDNF acutely modulates synaptic transmission and calcium signalling in developing cortical neurons. Cell Physiol Biochem 16:69 –76. Hellweg R, von Arnim CAF, Buchner M, Huber R, Riepe MW (2003): Neuroprotection and neuronal dysfunction upon repetitive inhibition of oxidative phosphorylation. Exp Neurol 183:346 –354. Jung RE, Haier RJ, Yeo RA, Rowland LM, Petropoulos H, Levine AS, et al.(2005): Sex differences in N-acetylaspartate correlates of general intelligence: An 1H-MRS study of normal human brain. Neuroimage 26:965– 972. Kadota T, Horinouchi T, Kuroda C (2001): Development and aging of the cerebrum: Assessment with proton MR spectroscopy. AJNR Am J Neuroradiol 22:128 –135. Karege F, Schwald M, Cisse M (2002): Postnatal developmental profile of brain-derived neurotrophic factor in rat brain and platelets. Neurosci Lett 328:261–264. Kim DS, Garwood M (2003): High-field magnetic resonance techniques for brain research. Curr Opin Neurobiol 13:612– 619. Kreis R (2004): Issues of spectral quality in clinical 1H-magnetic resonance spectroscopy and a gallery of artifacts. NMR Biomed 17:361–381. Lang UE, Hellweg R, Gallinat J (2005): Association of BDNF serum concentrations with central serotonergic activity: Evidence from auditory signal processing. Neuropsychopharmacology 30:1148 –1153. Levine ES, Kolb JE (2000): Brain-derived neurotrophic factor increases activity of NR2B-containing N-methyl-D-aspartate receptors in excised patches from hippocampal neurons. J Neurosci Res 62:357–362. Lommatzsch M, Zingler D, Schuhbaeck K, Schloetcke K, Zingler C, SchuffWerner P, et al.(2005): The impact of age, weight and gender on BDNF levels in human platelets and plasma. Neurobiol Aging 26:115–123. Lucetti C, Del Dotto P, Gambaccini G, Bernardini S, Bianchi MC, Tosetti M, et al.(2001): Proton magnetic resonance spectroscopy (1H-MRS) of motor cortex and basal ganglia in de novo Parkinson’s disease patients. Neurol Sci 22:69 –70. Mlynarik V, Gruber S, Moser E (2001): Proton T (1) and T (2) relaxation times of human brain metabolites at 3 Tesla. NMR Biomed 14:325–331. Nemoto K, Ohnishi T, Mori T, Moriguchi Y, Hashimoto R, Asada T, et al.(2006): The Val66Met polymorphism of the brain-derived neurotrophic factor gene affects age-related brain morphology. Neurosci Lett 397:25–29. Pan W, Banks WA, Fasold MB, Bluth J, Kastin AJ (1998): Transport of brainderived neurotrophic factor across the blood-brain barrier. Neuropharmacology 37:1553–1561.
BIOL PSYCHIATRY 2007;62:530 –535 535 Phillips HS, Hains JM, Laramee GR, Rosenthal A, Winslow JW (1990): Widespread expression of BDNF but not NT3 by target areas of basal forebrain cholinergic neurons. Science 250:290 –294. Poo MM (2001): Neurotrophins as synaptic modulators. Nat Rev Neurosci 2:24 –32. Porcellati G, Arienti G (1983): Metabolism of phosphoglycerides. In: A Lajtha, editor. Handbook of Neurochemistry, Metabolism in the Nervous System. New York: Plenum, 133–161. Rantamaki T, Knuuttila JE, Hokkanen ME, Castren E (2006): The effects of acute and long-term lithium treatments on trkB neurotrophin receptor activation in the mouse hippocampus and anterior cingulate cortex. Neuropharmacology 50:421– 427. Rex CS, Lauterborn JC, Lin CY, Kramar EA, Rogers GA, Gall CM, et al.(2006): Restoration of long-term potentiation in middle-aged hippocampus after induction of brain-derived neurotrophic factor. J Neurophysiol 96: 677– 685. Sager TN, Topp S, Torup L, Hanson LG, Egestad B, Moller A (2001): Evaluation of CA1 damage using single-voxel 1H-MRS and un-biased stereology: Can non-invasive measures of N-acetyl-asparate following global ischemia be used as a reliable measure of neuronal damage? Brain Res 892:166 –175. Schubert F, Gallinat J, Seifert F, Rinneberg H (2004): Glutamate concentrations in human brain using single voxel proton magnetic resonance spectroscopy at 3 Tesla. Neuroimage 21:1762–1771. Schulte-Herbruggen O, Chourbaji S, Müller H, Danker-Hopfe H, Brandwein C, Gass P, et al.(2006): Differential regulation of nerve growth factor and brain-derived neurotrophic factor in a mouse model of learned helplessness. Exp Neurol 202:404 – 409. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al.(1998): The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 59:22–33. Takahashi M, Shirakawa O, Toyooka K, Kitamura N, Hashimoto T, Maeda K, et al.(2000): Abnormal expression of brain-derived neurotrophic factor and its receptor in the corticolimbic system of schizophrenic patients. Mol Psychiatry 5:293–300. Terpstra M, Gruetter R (2004): 1H NMR detection of vitamin C in human brain in vivo. Magn Reson Med 51:225–229. Theberge J, Bartha R, Drost DJ, Menon RS, Malla A, Takhar J, et al.(2002): Glutamate and glutamine measured with 4.0 T proton MRS in nevertreated patients with schizophrenia and healthy volunteers. Am J Psychiatry 159:1944 –1946. Tkac I, Andersen P, Adriany G, Merkle H, Ugurbil K, Gruetter R (2001): In vivo 1H NMR spectroscopy of the human brain at 7 T. Magn Reson Med 46:451– 456. Trabesinger AH, Weber OM, Duc CO, Boesiger P (1999): Detection of glutathione in the human brain in vivo by means of double quantum coherence filtering. Magn Reson Med 42:283–289. Webster MJ, Herman MM, Kleinman JE, Shannon Weickert C (2006): BDNF and trkB mRNA expression in the hippocampus and temporal cortex during the human lifespan. Gene Expr Patterns 6:941–951. Weickert CS, Hyde TM, Lipska BK, Herman MM, Weinberger DR, Kleinman JE (2003): Reduced brain-derived neurotrophic factor in prefrontal cortex of patients with schizophrenia. Mol Psychiatry 8:592– 610. Yildiz-Yesiloglu A, Ankerst DP (2006): Neurochemical alterations of the brain in bipolar disorder and their implications for pathophysiology: A systematic review of the in vivo proton magnetic resonance spectroscopy findings. Prog Neuropsychopharmacol Biol Psychiatry 30:969 –995. Ziegenhorn AA, Schulte-Herbruggen O, Danker-Hopfe H, Malbranc M, Hartung HD, Anders D, et al.(2006): Serum neurotrophins-A study on the time course and influencing factors in a large old age sample. Neurobiol Aging Jul 29 [Epub ahead of print] Doi 10.1016/j.neurobiolaging. 2006.06.011. Zuccato C, Liber D, Ramos C, Tarditi A, Rigamonti D, Tartari M, et al.(2005): Progressive loss of BDNF in a mouse model of Huntington’s disease and rescue by BDNF delivery. Pharmacol Res 52:133–139.
www.sobp.org/journal
B
rain-derived neurotrophic factor (BDNF) signaling at synapses enhances neuronal plasticity and long-term potentiation (LTP), which represents a process of synaptic strengthening associated with learning and memory (Ernfors and Braham 2003; Figurov et al. 1996; Poo, 2001; Rex et al. 2006). Likewise, long-term developmental phenomena, such as neuronal survival, migration, and differentiation are mediated by BDNF through its ability to promote activity-dependent refinement of synaptic architecture (Baquet et al. 2004; Gorski et al. 2003; Poo 2001). The highest levels of BDNF are observed in hippocampus and cerebral cortex (Phillips et al. 1990), brain regions that are predominantly involved in higher cognitive functions (e.g., Gallinat et al. 2006). Expression of BDNF in these brain regions has been implicated in neurodevelopmental and neurodegenerative disorders, including Alzheimer’s disease, Huntington’s disease, and Down’s syndrome (Baquet et al. 2004; Bimonte-Nelson et al. 2003; Weickert et al. 2003; Zuccato et al. 2005), and abnormal BDNF concentrations in the anterior cingulate cortex (ACC) and hippocampus have been postulated as molecular substrate for pathological manifestations of schizophrenia (Takahashi et al. 2000). Because it has been described that intact BDNF in the peripheral circulation crosses the blood-brain barrier (Pan et al.
From the Department of Psychiatry and Psychotherapy (UEL, JG), Charité – University Medicine Berlin, Campus Mitte; Department of Psychiatry (RH), Charité—University Medicine Berlin, Campus Benjamin Franklin; and Physikalisch-Technische Bundesanstalt (FS, FS), Berlin, Germany. Address reprint requests to J. Gallinat, M.D., Ph.D., Charité Medicine Berlin, Clinic for Psychiatry and Psychotherapy, St. Hedwig Krankenhaus, Turmstrasse 21, 10559 Berlin, Germany; E-mail: [email protected]. Received October 12, 2006; revised January 3, 2007; accepted January 3, 2007.
0006-3223/07/$32.00 doi:10.1016/j.biopsych.2007.01.002
1998), BDNF blood concentrations might reflect central BDNF concentrations. However, an association of serum BDNF concentration with in vivo measured parameters of neuronal plasticity in humans has not been investigated so far. Although such a correlation would not prove a causal relationship between parameters, it would represent an argument for a cerebral role of peripheral measured BDNF. N-acetylaspartate (NAA) is a well established biochemical marker of neuronal damage, and its measurement in the acute phase of brain injury indicates the degree of neuronal loss and therefore reflects tissue metabolic dysfunction (Davie et al. 1997; Ebisu et al. 1994; Sager et al. 2001). Because cellular dysfunction can cause stronger reductions in NAA level than neuronal loss, NAA quantification has been suggested as a potential tool for visualizing the penumbra area in stroke patients (Demougeot et al. 2004). The NAA changes in mild cognitive impairment, Alzheimer’s disease (Ackl et al. 2005), and Parkinson=s disease (Lucetti et al. 2001) have been described. Apart from pathological conditions, cerebral NAA concentrations measured in healthy subjects have been associated with memory tasks (Gimenez et al. 2004) and intellectual performance and intelligence (Jung et al. 2005) as well as personality traits (Gallinat et al. 2005). Moreover, a use-dependent increase of cortical NAA has been described suggesting that this parameter might reflect neuronal adaption and plasticity in humans (Aydin et al. 2005). The technique of proton magnetic resonance spectroscopy (1H-MRS) permits the quantification of a number of brain metabolites noninvasively; with regard to sensitivity and spectral resolution (i.e., metabolite selectivity), it has benefited to a large extent from the increasing field strength of magnetic resonance (MR) scanners (e.g., Bartha et al. 2000; Kim and Garwood 2003; Tkac et al. 2001). Thus, using a field strength ⬎ 1.5-Tesla not only improves the signal-to-noise ratio for the commonly detected singlet resonances of NAA, choline-containing compounds (tCho), and creatine ⫹ phosphocreatine (tCr) but also permits, in principle, the estimation of the levels of other metabolites, such BIOL PSYCHIATRY 2007;62:530 –535 © 2007 Society of Biological Psychiatry
U.E. Lang et al. as neurotransmitters (Theberge et al. 2002), Vitamin C (Terpstra and Gruetter 2004), or reduced glutathione (Trabesinger et al. 1999). Because we have developed a methodology for the quantification of the aforementioned singlets together with Lglutamate at 3-Tesla (Gallinat et al. 2006; Schubert et al. 2004), we used this approach to test the hypothesis that BDNF serum concentration is associated with in vivo measured brain NAA as a marker of neuronal integrity in healthy subjects. In an exploratory manner the relationships of BDNF to other metabolic parameters (tCho, tCr, glutamate) were investigated. The present measurements were performed in regions displaying high tissue concentrations of BDNF, namely the ACC and the left hippocampus (Schulte-Herbruggen et al. 2006). It was hypothesized that NAA would show a positive correlation with BDNF serum concentration, because BDNF has been implicated in the structure and function of the adult brain (Bath and Lee 2006).
Methods and Materials Subjects The study was approved by the ethics committee of the Charité University Medicine Berlin. Thirty-six healthy subjects (age: 34.4 ⫾ 10.1 years, 17 men 33.8 ⫾ 8.9 years, 19 women 34.9 ⫾ 11.2 years) were recruited through newspaper advertisements. They were of German descent and gave written informed consent. All subjects were free of medical, neurological, and psychiatric disorders as determined by in-person interviews (Mini-International Neuropsychiatric Interview [M.I.N.I.]; Sheehan et al. 1998) performed by a psychiatrist. Subjects with a family history (first degree) of axis I disorder were excluded from participation. For further methodological details, see Gallinat et
BIOL PSYCHIATRY 2007;62:530 –535 531 al. (2002) and Lang et al. (2005). Blood for BDNF determination in serum was collected before the 1H-MRS investigation. MR Spectroscopy Magnetic resonance measurements were carried out on a 3-Tesla scanner (MEDSPEC 30/100; Bruker Biospin, Ettlingen, Germany) with a circularly polarized head coil. After automated global shim of the linear, xz, z2, and x2-y2 field components, T1-weighted images (modified driven equilibrium Fourier transform [MDEFT], echo time (TE) ⫽ 5.5 msec, repetition time (TR) ⫽ 23.4 msec, 64 contiguous slices of 2-mm thickness, 1-mm inplane (x-y) resolution) were acquired. The MR spectra were recorded from 2 ⫻ 3 ⫻ 2 cm3 voxels including the left hippocampus and from 2.5 ⫻ 4 ⫻ 2 cm3 voxels including the ACC of the volunteer brains (Figure 1). For metabolite quantification we closely followed an established method (Schubert et al. 2004). After manual shimming to water linewidths (full width at half maximum [FWHM]) of 7–9 Hz and 6 –7 Hz for the hippocampus and ACC voxels, respectively, and determination of the radiofrequency power needed for a 90° excitation pulse, calibration of water suppression was carried out, followed by acquisition of spectra with point resolved spectroscopy (PRESS). To obtain one MR spectrum, eight subspectra of 16 phase cycled scans each were recorded with TR ⫽ 3 sec and TE ⫽ 80 msec, giving 128 averages. Subsequently the water suppression pulses were switched off to acquire a waterunsuppressed spectrum (n ⫽ 8). Before further processing, the eight individual metabolite subspectra were corrected for eddy currents with the water-unsuppressed spectrum and automatically frequency aligned to correct for frequency shifts during the
Figure 1. Voxel positions shown on typical brain modified driven equilibrium Fourier transform (MDEFT) images; anterior cingulate cortex (left), hippocampus (right).
www.sobp.org/journal
532 BIOL PSYCHIATRY 2007;62:530 –535
Figure 2. Point resolved spectroscopy (PRESS) spectra (repetition time [TR] ⫽ 3 sec, and echo time [TE] ⫽ 80 msec) acquired from voxels containing the anterior cingulate cortex (ACC) (top) and left hippocampus (bottom). tCr, total creatine; tCho, total choline; NAA, N-acetylaspartate; Glu, glutamate. Note that, apart from phasing, no postprocessing was carried out.
scan time (of more than 6 min), caused by involuntary subject motion and system instabilities. Examples of resulting spectra— then of 128 averages—are shown in Figure 2. The spectral linewidth (FWHM) values of the metabolite resonances are estimated by the used fitting procedure (described in following text); mean FWHM (SD) (in Hz) were 4.9 (.6) for the ACC voxel and 6.6 (.9) for the hippocampus voxel and thus below the minimum acceptable linewidth for high-precision MR spectroscopy suggested recently (Kreis 2004). Spectra were quantified with a program package that relies on a time domain-frequency domain fitting procedure and involves inclusion of phantom basis spectra and prior knowledge into the fit (Elster et al. 2005; Schubert et al. 2004). At the chosen TE of 80 msec the baseline was not entirely flat. To account for baseline features, the fitting procedure routinely includes estimation of the baseline nonparametrically by regularization. In the present spectra, total choline (tCho), total creatine (tCr), NAA, glutamate (Glu), and glutamine resonances were fitted by inclusion of phantom spectra for the latter three and imposing constant frequency differences for glutamate, glutamine, and NAA and equal linewidths as prior knowledge. Extensive tests yielded mean uncertainties (corresponding to Cramér-Rao lower bounds with added uncertainties from background modeling) for the fitting of NAA of 2.4% and 2.7%, tCho 3.7% and 4.0%, tCr 3.0% and 3.1%, and Glu 11.0% and 13.2% for the ACC and the hippocampus voxel, respectively. Metabolite amplitudes returned by the fitting procedure were corrected for different coil loading by the phantoms and the individual subject’s head (principle of reciprocity [Danielsen and Henriksen 1994]) and for relaxation effects, assuming no differences in relaxation behavior between the www.sobp.org/journal
U.E. Lang et al. subjects. Transverse relaxation times were determined from three healthy volunteers with TE of 50, 80, 135, 250, and 330 msec. The results (in msec, SD in parentheses) for Glu(C4), NAA, tCr, and tCho were 194 (37), 278 (31), 179 (9), and 282 (45); and 171 (22), 267 (15), 198 (31), and 291 (13) for the ACC voxel and for the hippocampus voxel, respectively. It should be noted that, although perfectly applicable for the correction, the transverse relaxation time of Glu is only a relative one, because it was determined by fitting of the in vivo data to phantom spectra acquired at the same TE. As T1 for NAA, tCr, and tCho, the following mean values determined from two healthy volunteers with TE ⫽ 30 msec and TR ⫽ .5, 1, 2, 3, 5, and 7 s were employed in the analysis: 1.48 sec; 1.21 sec; and 1.44 sec. For Glu, no correction for longitudinal relaxation effects was carried out. The deviation caused by the T1 effect at TR ⫽ 3 sec, and an assumed T1 of glutamate at 3-T of approximately 1.2 sec (Mlynarik et al. 2001) was largely compensated by the deviation estimated for the aqueous, buffered glutamate phantom, where TR was set at 5 sec and T1 was determined to be 1.47 sec. Thus, glutamate might at worst be systematically underestimated by about 5%, owing to unaccounted T1 effects, which would lead to uniform scaling of the calculated concentrations. Although necessary for a stable Glu quantification, the inclusion of glutamine in the fit gave only an estimate of its concentration, the uncertainty of which was too large for further consideration (Schubert et al. 2004). The in vivo concentrations were corrected for the content of cerebrospinal fluid (CSF) of the voxels studied, as derived from segmentation of the T1-weighted images with spm99 (Ashburner and Friston 1997). Providing the highest retest reliability (Schubert et al. 2004), a pixel was classified depending on which spm99 tissue classification had the greatest probability. Serum aliquots for neurotrophin measurements were stored at ⫺70° C before batch processing. Levels of endogenous BDNF were measured in diluted serum samples by a highly sensitive and specific fluorometric two-site enzyme-linked immunosorbent assay with commercial kits, in principle according to the manufacturer’s instructions (Promega, Mannheim, Germany) but in an improved, fluorometric form described previously (Hellweg et al. 2003, Ziegenhorn et al. 2006). Statistical Analyses To study the relationships between BDNF serum concentrations and brain metabolites, linear regression analysis was performed with the independent factors brain metabolites (NAA, tCho, tCr, Glu), age, and gender and the dependent variable BDNF serum concentration. Linear regression analyses were separately computed for hippocampal and ACC metabolite concentrations. For correlation analyses, Pearson correlation coefficients were computed. All tests were performed at a level of significance of p ⫽ .05.
Results The mean BDNF serum concentration amounted to 15.8 ⫾ 8.4 ng/mL. The absolute metabolite concentrations (in mmol/L) in the ACC were 13.6 ⫾ 1.2 for NAA, 2.14 ⫾ .25 for total choline (tCho), 9.4 ⫾ 1.0 for total creatine (tCr), and 11.5 ⫾ 1.6 for Glu. In the hippocampus the concentrations (in mmol/L) were 11.6 ⫾ 1.0 for NAA, 2.21 ⫾ .26 for tCho, 9.6 ⫾ .6 for tCr, and 10.9 ⫾ 1.2 for Glu. The linear regression analysis revealed that BDNF serum concentration was predicted by the concentrations of NAA (T ⫽ 2.193, p ⫽ .037) and, as a statistical trend, tCho (T ⫽ 1.997, p ⫽ .055) measured in the ACC. The result with regard to NAA
U.E. Lang et al.
BIOL PSYCHIATRY 2007;62:530 –535 533
Figure 3. Scatterplots of NAA concentrations measured in the ACC versus serum brain-derived neurotrophic factor (BDNF) concentrations (A) and their residuals after linear correction for age and gender (B). Significant positive correlations were observed in both analyses (r ⫽ .396, p ⫽ .017 and r ⫽ .448, p ⫽ .008, respectively). Other abbreviations as in Figure 2.
remained significant (T ⫽ 2.111, p ⫽ .044) when the regression analysis was repeated with the additional factor gray matter content of the voxel (T ⫽ ⫺.190, p ⫽ .851). No other metabolite concentration (i.e., tCr or Glu) was a significant predictor, nor were age or gender. The linear regression analysis of BDNF concentration with the hippocampal metabolite concentrations as predictor variables did not reveal significant effects. This was also the case when the gray matter content of the voxel was included in the analysis. A correlation analysis was carried out for the brain metabolite significantly predicting the BDNF concentration. Pearson correlation analysis showed a significant positive correlation between BDNF concentration and the concentration of NAA in the ACC (r ⫽ .396, p ⫽ .017; Figure 3A). This association was also significant when the correlation analysis was controlled for age and gender (r ⫽ .448, p ⫽ .008, partial correlation; Figure 3B).
Discussion We found a positive correlation between cortical NAA concentration and BDNF serum concentration in healthy human subjects. No significant relationships between BDNF serum concentrations and other cortical or any hippocampal brain metabolites were observed. The results indicate that neuronal integrity and vitality of a cortical region like the ACC, as reflected by a high concentration of NAA, might be reflected by high serum concentrations of BDNF. Brain-derived neurotrophic factor (BDNF), like other neurotrophins, has long-term effects on neuronal survival and differentiation (Poo 2001). Thus prefrontally targeted BDNF knock-out mice have been shown to display reductions in cortical volume during adulthood (Baquet et al. 2004); an age-related volume reduction was found in the dorsolateral prefrontal cortices in humans carrying the Met-allele of the BDNF gene (Nemoto et al. 2006). These findings indicate a role of BDNF for the morphology and integrity of cortical brain areas that is in line with the positive correlation between serum BDNF and cortical NAA concentrations observed in our study. Further evidence indicates that the cortical level of NAA also reflects brain function (Bertolino et al. 1999) as well as behavioral aspects or personality traits
(see opening text of this report). For instance, in an in vivo 1H-MRS study an association was found between cortical NAA (ACC) and the level of cognitive interference in the Stroop Color-Word task (Grachev et al. 2001), a paradigm known to be dependent on ACC function (Bench et al. 1993). Grachev et al. suggested their results to be a correlate of a dynamic process of neuronal reorganization in the ACC resulting from use-dependent activity. This view is compatible with a recently reported use-dependent cortical NAA increase in humans (Aydin et al. 2005) and with findings showing neurotrophins to rapidly increase spontaneous firing rate and synaptic transmission in cortical neurons (He et al. 2005). The BDNF signaling at synapses enhances activity-dependent long-term potentiation (Ernfors and Bramham 2003; Figurov et al. 1996; Levine and Kolb 2000; Poo 2001), a process relevant for long-term plasticity in the ACC. In this context, the present results might indicate that high cortical NAA concentrations reflect ongoing neuronal plasticity mediated by BDNF. Interestingly, the majority of reports describe an association of BDNF with functional (Egan et al. 2003; Hariri et al. 2003) and structural aspects (Bueller et al. 2006) of the hippocampus, a brain region with a high expression of BDNF protein. In the present study, no association between BDNF serum concentration and hippocampal neurochemistry was observed, even when the analysis was controlled for the gray matter content of the hippocampal voxel. There could be several reasons for this result: for instance, the relatively large voxel size, including regions other than the hippocampus, and a slightly higher uncertainty of metabolite quantification compared with that in the ACC (Schubert et al. 2004). Another confounding factor might be the reported effect of a missense variation of the BDNF gene (Val66Met substitution) on the level of NAA in the hippocampus of healthy volunteers (Egan et al. 2003). However, the most important reason for a lacking correlation between BDNF and hippocampal metabolites might be the relatively small size of the hippocampus compared with the cerebral cortex in relation to whole brain volume. Given that cerebral BDNF crosses the blood-brain barrier (Pan et al. 1998), it is reasonable to assume that serum BDNF concentrations are associated more closely www.sobp.org/journal
534 BIOL PSYCHIATRY 2007;62:530 –535 with neocortical than with hippocampal BDNF levels. This assumption would be in line with the finding in an animal experiment that BDNF in serum is correlated with BDNF expression in cortical but not hippocampal region (Karege et al. 2002). Other evidence indicates that BDNF as well as NAA concentrations are modulated by pharmacological intervention more dynamically in the ACC (or cortical regions) than in the hippocampus (Rantamaki et al. 2006; Yildiz-Yesiloglu and Ankerst 2006). Dynamic BDNF regulation might play a role for the observed positive correlation between cortical NAA and serum BDNF. In a BDNF mutant mouse model the hypothesis has been worked out that BDNF supports the maintenance of cortical neuron size and dendrite structure rather than the initial development of these features (Gorski et al. 2003). For this reason we assume permanent BDNF support to be required at least in cortical areas to maintain neuronal functioning and suggest NAA to be a marker of this process. This model is also in line with an age-dependent decline of serum and cortical BDNF levels (Webster et al. 2006; Ziegenhorn et al. 2006) as well as cerebral NAA (Schubert et al. 2004). Of note, BDNF concentration in human hippocampus does not change significantly over the life span (Webster et al. 2006). However, other arguments contradict the role of NAA as a marker of neuronal plasticity. For instance, the concentration of NAA in infants during development is lower compared with adults (Kadota et al. 2001). Furthermore, several biological conditions, like menstrual cycle or body weight, have an effect on BDNF concentrations in the peripheral blood (Lommatzsch et al. 2005), whereas cerebral levels of NAA are largely unaffected by these factors (Epperson et al. 2005). However, the effect sizes of the mentioned biological factors on BDNF concentration are rather small, which might not completely exclude a relationship between BDNF and the cerebral metabolite NAA. In principle, the use of the current datasets in previously published papers (Gallinat et al. 2005, 2006, 2007; Schubert et al. 2004) increases the risk of false positive results, which have to be taken into account when interpreting the actual results. A trend was observed for the concentrations of cortical tCho to increase with increased serum BDNF. Although not significant, this is compatible with a dynamic maintenance of neurons driven by cortical BDNF, because tCho fluctuates when cellular membranes are degraded or rapidly synthesized (Dawson 1985; Porcellati and Arienti 1983). We did not find a significant correlation between BDNF serum concentration and cerebral glutamate, although both components have previously been shown to interact in several ways (Bustos et al. 2004). However, the precision of cerebral glutamate measurement is lower than that for NAA (Schubert et al. 2004), which might hinder the detection of a relationship with serum BDNF. In conclusion, we detected for the first time an association between serum BDNF protein concentration and cortical NAA concentration in the living human brain. The results are in line with the formerly described role of neurotrophins as regulatory factors mediating the differentiation and survival of neurons (Poo 2001) as reflected by the level of NAA.
Support from the German Federal Ministry of Education and Research (BMBF, Project Berlin Neuroimaging Center, No. 01G00208) is gratefully acknowledged. Drs. Lang and Hellweg contributed equally to this paper. www.sobp.org/journal
U.E. Lang et al. Ackl N, Ising M, Schreiber YA, Atiya M, Sonntag A, Auer DP (2005): Hippocampal metabolic abnormalities in mild cognitive impairment and Alzheimer’s disease. Neurosci Lett 384:23–28. Ashburner J, Friston K (1997): Multimodal image coregistration and partitioning—a unified framework. Neuroimage 6:209 –217. Aydin K, Ciftci K, Terzibasioglu E, Ozkan M, Demirtas A, Sencer S, et al.(2005): Quantitative proton MR spectroscopic findings of cortical reorganization in the auditory cortex of musicians. AJNR Am J Neuroradiol 26:128 – 136. Baquet ZC, Gorski JA, Jones KR (2004): Early striatal dendrite deficits followed by neuron loss with advanced age in the absence of anterograde cortical brain-derived neurotrophic factor. J Neurosci 24:4250 – 4258. Bartha R, Drost DJ, Menon RS, Williamson PC (2000): Comparison of the quantification precision of human short echo time (1)H spectroscopy at 1.5 and 4.0 Tesla. Magn Reson Med 44:185–192. Bath KG, Lee FS (2006): Variant BDNF (Val66Met) impact on brain structure and function. Cogn Affect Behav Neurosci 6:79 – 85. Bench CJ, Frith CD, Grasby PM, Friston KJ, Paulesu E, Frackowiak RS, et al.(1993): Investigations of the functional anatomy of attention using the Stroop test. Neuropsychologia 31:907–922. Bertolino A, Knable MB, Saunders RC, Callicott JH, Kolachana B, Mattay VS, et al.(1999): The relationship between dorsolateral prefrontal N-acetylaspartate measures and striatal dopamine activity in schizophrenia. Biol Psychiatry 45:660 – 667. Bimonte-Nelson HA, Hunter CL, Nelson ME, Granholm AC (2003): Frontal cortex BDNF levels correlate with working memory in an animal model of Down syndrome. Behav Brain Res 139:47–57. Bueller JA, Aftab M, Sen S, Gomez-Hassan D, Burmeister M, Zubieta JK (2006): BDNF Val(66)Met al.ele is associated with reduced hippocampal volume in healthy subjects. Biol Psychiatry 59:812– 815. Bustos G, Abarca J, Campusano J, Bustos V, Noriega V, Aliaga E (2004): Functional interactions between somatodendritic dopamine release, glutamate receptors and brain-derived neurotrophic factor expression in mesencephalic structures of the brain. Brain Res Brain Res Rev 47:126 – 144. Danielsen ER, Henriksen O (1994): Absolute quantitative proton NMR spectroscopy based on the amplitude of the local water suppression pulse. Quantification of brain water and metabolites. NMR Biomed 7:311–318. Davie CA, Barker GJ, Thompson AJ, Tofts PS, McDonald WI, Miller DH (1997): 1H magnetic resonance spectroscopy of chronic cerebral white matter lesions and normal appearing white matter in multiple sclerosis. J Neurol Neurosurg Psychiatry 63:736 –742. Dawson RMC (1985): Enzymatic pathways of phospholipid metabolism in the nervous system. In: Eichberg J, editor. Phospholipids in Nervous Tissues. New York: Wiley, 45–78. Demougeot C, Marie C, Giroud M, Beley A (2004): N-acetylaspartate: A literature review of animal research on brain ischaemia. J Neurochem 90: 776 –783. Ebisu T, Rooney WD, Graham SH, Weiner MW, Maudsley AA (1994): N-acetylaspartate as an in vivo marker of neuronal viability in kainate-induced status epilepticus: 1H magnetic resonance spectroscopic imaging. J Cereb Blood Flow Metab 14:373–382. Egan MF, Kojima M, Callicott JH, Goldberg TE, Kolachana BS, Bertolino A, et al.(2003): The BDNF val66met polymorphism affects activity-dependent secretion of BDNF and human memory and hippocampal function. Cell 112:257–269. Elster C, Schubert F, Link A, Walzel M, Seifert F, Rinneberg H (2005): Quantitative magnetic resonance spectroscopy: Semi-parametric modeling and determination of uncertainties. Magn Reson Med 53:1288 –1296. Epperson CN, O’Malley S, Czarkowski KA, Gueorguieva R, Jatlow P, Sanacora G, et al.(2005): Sex, GABA, and nicotine: The impact of smoking on cortical GABA levels across the menstrual cycle as measured with proton magnetic resonance spectroscopy. Biol Psychiatry 57:44 – 48. Ernfors P, Bramham CR (2003): The coupling of a trkB tyrosine residue to LTP. Trends Neurosci 26:171–173. Figurov A, Pozzo-Miller LD, Olafsson P, Wang T, Lu B (1996): Regulation of synaptic responses to high-frequency stimulation and LTP by neurotrophins in the hippocampus. Nature 381:706 –709. Gallinat J, Kunz D, Lang UE, Neu P, Kassim N, Kienast T, et al.(2007): Association between cerebral glutamate and human behaviour: The sensation seeking personality trait. Neuroimage 34:671– 678.
U.E. Lang et al. Gallinat J, Kunz D, Senkowski D, Kienast T, Seifert F, Schubert F, et al.(2006): Hippocampal glutamate concentration predicts cerebral theta oscillations in human memory utilization. Psychopharmacology 187:103–111. Gallinat J, Mulert C, Bajbouj M, Herrmann WM, Schunter J, Senkowski D, et al.(2002): Frontal and temporal dysfunction of auditory stimulus processing in schizophrenia. Neuroimage 17:110 –127. Gallinat J, Ströhle A, Lang UE, Bajbouj M, Kalus P, Montag C, et al.(2005): Association of human hippocampal neurochemistry, serotonin transporter genetic variation, and anxiety. Neuroimage 26:123–131. Gimenez M, Junque C, Narberhaus A, Caldu X, Segarra D, Vendrell P, et al.(2004): Medial temporal MR spectroscopy is related to memory performance in normal adolescent subjects. Neuroreport 15:703–707. Gorski JA, Zeiler SR, Tamowski S, Jones KR (2003): Brain-derived neurotrophic factor is required for the maintenance of cortical dendrites. J Neurosci 23:6856 – 6865. Grachev ID, Kumar R, Ramachandran TS, Szeverenyi NM (2001): Cognitive interference is associated with neuronal marker N-acetyl aspartate in the anterior cingulate cortex: An in vivo (1)H-MRS study of the Stroop ColorWord task. Mol Psychiatry 6:529 –539. Hariri AR, Goldberg TE, Mattay VS, Kolachana BS, Callicott JH, Egan MF, et al.(2003): Brain-derived neurotrophic factor val66met polymorphism affects human memory related hippocampal activity and predicts memory performance. J Neurosci 23:6690 – 6694. He J, Gong H, Luo Q (2005): BDNF acutely modulates synaptic transmission and calcium signalling in developing cortical neurons. Cell Physiol Biochem 16:69 –76. Hellweg R, von Arnim CAF, Buchner M, Huber R, Riepe MW (2003): Neuroprotection and neuronal dysfunction upon repetitive inhibition of oxidative phosphorylation. Exp Neurol 183:346 –354. Jung RE, Haier RJ, Yeo RA, Rowland LM, Petropoulos H, Levine AS, et al.(2005): Sex differences in N-acetylaspartate correlates of general intelligence: An 1H-MRS study of normal human brain. Neuroimage 26:965– 972. Kadota T, Horinouchi T, Kuroda C (2001): Development and aging of the cerebrum: Assessment with proton MR spectroscopy. AJNR Am J Neuroradiol 22:128 –135. Karege F, Schwald M, Cisse M (2002): Postnatal developmental profile of brain-derived neurotrophic factor in rat brain and platelets. Neurosci Lett 328:261–264. Kim DS, Garwood M (2003): High-field magnetic resonance techniques for brain research. Curr Opin Neurobiol 13:612– 619. Kreis R (2004): Issues of spectral quality in clinical 1H-magnetic resonance spectroscopy and a gallery of artifacts. NMR Biomed 17:361–381. Lang UE, Hellweg R, Gallinat J (2005): Association of BDNF serum concentrations with central serotonergic activity: Evidence from auditory signal processing. Neuropsychopharmacology 30:1148 –1153. Levine ES, Kolb JE (2000): Brain-derived neurotrophic factor increases activity of NR2B-containing N-methyl-D-aspartate receptors in excised patches from hippocampal neurons. J Neurosci Res 62:357–362. Lommatzsch M, Zingler D, Schuhbaeck K, Schloetcke K, Zingler C, SchuffWerner P, et al.(2005): The impact of age, weight and gender on BDNF levels in human platelets and plasma. Neurobiol Aging 26:115–123. Lucetti C, Del Dotto P, Gambaccini G, Bernardini S, Bianchi MC, Tosetti M, et al.(2001): Proton magnetic resonance spectroscopy (1H-MRS) of motor cortex and basal ganglia in de novo Parkinson’s disease patients. Neurol Sci 22:69 –70. Mlynarik V, Gruber S, Moser E (2001): Proton T (1) and T (2) relaxation times of human brain metabolites at 3 Tesla. NMR Biomed 14:325–331. Nemoto K, Ohnishi T, Mori T, Moriguchi Y, Hashimoto R, Asada T, et al.(2006): The Val66Met polymorphism of the brain-derived neurotrophic factor gene affects age-related brain morphology. Neurosci Lett 397:25–29. Pan W, Banks WA, Fasold MB, Bluth J, Kastin AJ (1998): Transport of brainderived neurotrophic factor across the blood-brain barrier. Neuropharmacology 37:1553–1561.
BIOL PSYCHIATRY 2007;62:530 –535 535 Phillips HS, Hains JM, Laramee GR, Rosenthal A, Winslow JW (1990): Widespread expression of BDNF but not NT3 by target areas of basal forebrain cholinergic neurons. Science 250:290 –294. Poo MM (2001): Neurotrophins as synaptic modulators. Nat Rev Neurosci 2:24 –32. Porcellati G, Arienti G (1983): Metabolism of phosphoglycerides. In: A Lajtha, editor. Handbook of Neurochemistry, Metabolism in the Nervous System. New York: Plenum, 133–161. Rantamaki T, Knuuttila JE, Hokkanen ME, Castren E (2006): The effects of acute and long-term lithium treatments on trkB neurotrophin receptor activation in the mouse hippocampus and anterior cingulate cortex. Neuropharmacology 50:421– 427. Rex CS, Lauterborn JC, Lin CY, Kramar EA, Rogers GA, Gall CM, et al.(2006): Restoration of long-term potentiation in middle-aged hippocampus after induction of brain-derived neurotrophic factor. J Neurophysiol 96: 677– 685. Sager TN, Topp S, Torup L, Hanson LG, Egestad B, Moller A (2001): Evaluation of CA1 damage using single-voxel 1H-MRS and un-biased stereology: Can non-invasive measures of N-acetyl-asparate following global ischemia be used as a reliable measure of neuronal damage? Brain Res 892:166 –175. Schubert F, Gallinat J, Seifert F, Rinneberg H (2004): Glutamate concentrations in human brain using single voxel proton magnetic resonance spectroscopy at 3 Tesla. Neuroimage 21:1762–1771. Schulte-Herbruggen O, Chourbaji S, Müller H, Danker-Hopfe H, Brandwein C, Gass P, et al.(2006): Differential regulation of nerve growth factor and brain-derived neurotrophic factor in a mouse model of learned helplessness. Exp Neurol 202:404 – 409. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al.(1998): The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 59:22–33. Takahashi M, Shirakawa O, Toyooka K, Kitamura N, Hashimoto T, Maeda K, et al.(2000): Abnormal expression of brain-derived neurotrophic factor and its receptor in the corticolimbic system of schizophrenic patients. Mol Psychiatry 5:293–300. Terpstra M, Gruetter R (2004): 1H NMR detection of vitamin C in human brain in vivo. Magn Reson Med 51:225–229. Theberge J, Bartha R, Drost DJ, Menon RS, Malla A, Takhar J, et al.(2002): Glutamate and glutamine measured with 4.0 T proton MRS in nevertreated patients with schizophrenia and healthy volunteers. Am J Psychiatry 159:1944 –1946. Tkac I, Andersen P, Adriany G, Merkle H, Ugurbil K, Gruetter R (2001): In vivo 1H NMR spectroscopy of the human brain at 7 T. Magn Reson Med 46:451– 456. Trabesinger AH, Weber OM, Duc CO, Boesiger P (1999): Detection of glutathione in the human brain in vivo by means of double quantum coherence filtering. Magn Reson Med 42:283–289. Webster MJ, Herman MM, Kleinman JE, Shannon Weickert C (2006): BDNF and trkB mRNA expression in the hippocampus and temporal cortex during the human lifespan. Gene Expr Patterns 6:941–951. Weickert CS, Hyde TM, Lipska BK, Herman MM, Weinberger DR, Kleinman JE (2003): Reduced brain-derived neurotrophic factor in prefrontal cortex of patients with schizophrenia. Mol Psychiatry 8:592– 610. Yildiz-Yesiloglu A, Ankerst DP (2006): Neurochemical alterations of the brain in bipolar disorder and their implications for pathophysiology: A systematic review of the in vivo proton magnetic resonance spectroscopy findings. Prog Neuropsychopharmacol Biol Psychiatry 30:969 –995. Ziegenhorn AA, Schulte-Herbruggen O, Danker-Hopfe H, Malbranc M, Hartung HD, Anders D, et al.(2006): Serum neurotrophins-A study on the time course and influencing factors in a large old age sample. Neurobiol Aging Jul 29 [Epub ahead of print] Doi 10.1016/j.neurobiolaging. 2006.06.011. Zuccato C, Liber D, Ramos C, Tarditi A, Rigamonti D, Tartari M, et al.(2005): Progressive loss of BDNF in a mouse model of Huntington’s disease and rescue by BDNF delivery. Pharmacol Res 52:133–139.
www.sobp.org/journal