Thalamic neurochemical abnormalities in individuals with prodromal symptoms of schizophrenia – relationship to auditory event-related potentials
James M Stone,a,b*
Elvira Bramon,b
Astrid Pauls,b
Alex Sumich,b,c
Philip K McGuireb
aDivision of Experimental Medicine
Imperial College London,
Burlington Danes Building,
Hammersmith Hospital,
Du Cane Road,
W12 0NN
bDepartment of Psychological Medicine and Psychiatry
King’s College London Institute of Psychiatry
De Crespigny Park
London
SE5 8AF
cDivision of Psychology
School of Social Sciences,
Nottingham Trent University
Burton Street
Nottingham
NG1 4BU
*corresponding author
email:
tel: 020 7848 0915
fax: 020 7848 0976
Type: Short communication
Word count: Abstract: 50 Main Text: 1493
Keywords: schizophrenia, psychosis, EEG, MMN, P300, Glutamate, Glutamine, NAA, Magnetic Resonance Spectroscopy. MRS
Thalamic neurochemical abnormalities may underlie psychotic symptoms and auditory event-related potential (ERP) abnormalities in schizophrenia. We investigated this hypothesis in subjects at risk of psychosis using MRS and EEG. Reduced thalamic glutamate plus glutamine and NAA levels were associated with abnormal frontal ERPs, supporting a thalamic basis for filtering impairments.
Individuals with an at risk mental state (ARMS) experience subthreshold psychotic symptoms and are at greatly increased risk of developing schizophrenia (Phillips et al. 2000). We recently reported ARMS subjects have neurochemical abnormalities in left thalamus, detected using proton magnetic resonance spectroscopy (1H-MRS) (Stone et al. 2009), includingreductions in thalamic glutamate (Glu), glutamate plus glutamine (Glx), andN-acetyl aspartate (NAA), markers of glutamatergic neurotransmission and neuronal integrity.
Thalamic abnormalities have been suggested to be of key importance inthe development of psychosis, withimpaired sensory filtering leading to cognitive deficits and thought disorder(Andreasen 1999). There is also accumulating evidence that the psychotogenic effects of NMDA receptor antagonists may occur through action in thalamus, as opposed to direct effects on cortical brain regions (Sharp et al. 2001).
Two electrophysiological measures – mismatch negativity (MMN) and P300 amplitude –are reduced in schizophrenia and in ARMS subjects (Bramon et al. 2004; Bramon et al. 2008; Sumich et al. 2008). Theyare putative indices of deviance detection and may reflect mechanisms involved in filtering relevant information into consciousness (Friedman et al. 2001). They have also been shown to be strongly influenced by thalamic functions (Klostermann et al. 2006). Both MMN and P300 are evoked in response the presentation of a deviant stimulus (one that differs in one or more physical properties from a preceding series of standard stimuli). MMN amplitude is dependent on the extent of this physical deviancy, whereasP300 relies on behavioural salience: frontal P300 amplitude increases with the capture of attention by novel, startling or contextually meaningful stimuli (Friedman et al. 2001).
MMN amplitude has been found to be dependent on glutamate function (Umbricht et al. 2000), whereas P300 amplitude in schizophrenia has been hypothesised to rely on neuronal integrity, as indexed by NAA levels (Molina et al. 2002).Based on these findings, we hypothesised that in ARMS subjects, thalamic NAA, glutamate and glutamine would be associated with frontal P300 and MMN amplitude.
Methods
Data from ARMS subjects who took part in previous 1H-MRS and EEG studies (Bramon et al. 2008; Stone et al. 2009), including previously unreported MMN data, were collated. A total of 11 ARMS subjects (5 female, mean(SD) age 27(4)) had data for both 1H-MRS and MMN, and 8 (5 female, mean(SD) age 29(5)) for 1H-MRS and P300, representing a subset of the total of 27 ARMS subjects who had MRS data, and 35 with EEG data. The mean(SD) time between MRS and EEG measures was 645(475) days. One subject with P300 data, and onesubject with MMN data had previous antipsychotic exposure
1H-MRS data acquisition
Proton MRS data were acquired on a General Electric (Milwaukee, USA) HDx 3T MR scanner. A whole brain 3D orthogonal coronal inversion recovery prepared spoiled gradient echo (IR-SPGR) scan was used for voxel localization and volume correction. A15 x 20 x 20 mm left thalamus ROI was defined at the point in the coronal slices where the thalamus was widest, using sagittal and coronal localizers to ensure that the ROI was clear of CSF contamination. Auto-prescan was performed before each scan, aiming for a line-width of less than 11Hz and optimization of shimming and water suppression. A single spectrum was then acquired using conventional Point Resolved Spectroscopy (PRESS) acquisition parameters (TE=30ms, TR=3000ms, 96 averages).
1H-MRS spectrum analysis
All spectra were analyzed, generating water-scaled metabolite levels, using LCModel version 6.1-4F (Provencher 1993). The IR-SPGR images were segmented using statistical parametric mapping software (SPM2 Wellcome Department of Imaging Neurosciences, University College London, United Kingdom) to allow correction of the spectroscopy results for partial volume CSF contamination. Metabolite levels were divided by the brain tissue (grey plus white matter) content of the voxel in each subject. Poorly fitted metabolite peaks (Cramer-Rao minimum variance bounds >20%) were excluded from further analysis.
EEG acquisition
EEG data were collected from 64 scalp sites according to the 10/20 International System and were grounded at FPZ using sintered electrodes in a cap. Bilateral mastoids served as reference and vertical and horizontal electro-oculographs monitored eye movements. Data were continuously digitised at 1000 Hz with a 0.05 to 100 Hz band-pass filter (24 dB/octave roll-off). Impedances were kept below 5 KOhms. A Neuroscan linear regression procedure was employed to minimise ocular artifacts (Semlitsch et al. 1986).
P300 paradigm
The P300 was elicited using an auditory oddball paradigm (Bramon et al. 2008). Stimuli were four hundred 80 dB tones, with a 2 (± 0.2) seconds inter-stimulus interval presented through bilateral intra-aural earphones. 80% of the tones were ‘non-targets’ of 1000 Hz and 20% were ‘targets’ of 1500 Hz in a random sequence. Subjects were instructed to keep their eyes open and press a button in response to target tones only.
Continuous EEG was epoched (-100 to 700 ms), baseline corrected using the pre-stimulus interval (-100 to 0 ms), band-pass filtered 0.05 to 45 Hz and averaged for targets and non-targets separately (Frangou et al. 1997; Bramon et al. 2008). The P300 was defined as a positive waveform generated by the target tones and peaking between 280-500 ms post-stimulus. P300 peak amplitudes and latencies were measured using a computer algorithm, which made the process blind to group affiliation. The peak amplitude values obtained at three frontal sites (FZ, F3 and F4) were used for statistical analyses.
MMN paradigm
Stimuli were 1200 80-dB, 1000-Hz tones, with a 0.3 seconds inter-stimulus interval. These were presented as three blocks of 400 stimuli through bilateral intra-aural earphones. Eighty-five percent of the tones were ‘standards’ (25-ms duration, 5-ms rise/fall time) and 15% were ‘deviants’ (50-ms duration, 5-ms rise/fall time). Subjects were sitting comfortably in an armchair and were instructed to keep their eyes open and disregard the sounds presented to them.
The continuous EEG recording was epoched from -100 to 300 ms post-stimulus, baseline corrected using the pre-stimulus interval, band-pass filtered 0.05 to 45 Hz and averaged separately for the standard and deviant tones. Mismatch negativity (MMN) was defined as the difference between the deviant and standard event-related potentials. The peak MMN 50 to 200 ms post-stimulus was identified by a computer algorithm, which made the process blind to group status. The peak amplitude values obtained at FZ, F3 and F4 were used for statistical analyses.
Statistical analysis
Stepwise multiple regression was used to test the relationship between left thalamic biochemistry (NAA, Glu and Glx), entered as dependent variables in two separate analyses, and ERP amplitudes (MMN and P300 from each of the 3 frontal sites) as explanatory variables. Bonferroni correction of p values was undertaken to account for the 3 metabolite measures studied.
Results
Mean(SD) left thalamic Glx (8.46(2.04)) and NAA (10.84(0.64)) levels were representative(within 1%) of previously reported values in ARMS subjects from this dataset, which had been found to be significantly lower than control values (Stone et al. 2009).
Left thalamic Glx levels were predicted by a model including MMN from left (F3) and midline (FZ) sites (beta = 2.3281, -1.5477; SE = 0.6917, 0.6610, t = 3.366, -2.341 respectively; F2,8=6.52; p=0.02; Bonferroni corrected p=0.06) (Fig 1). Post-hoc Pearson’s correlation analysis revealed that lower MMN amplitude at F3 was related to lower Glx levels (df=9 r=0.6, p=0.05). Lower NAA in left thalamus was associated with reduced right frontal (F4) P300 amplitude (beta = 0.19259, SE = 0.04204, t = 4.581, F1,6=20.98, r=0.88, p=0.004; Bonferroni corrected p=0.01) (Fig 2). Post-hoc testing revealed left thalamic NAA also correlated significantly with P300 amplitude at F3 and Fz (df=6, r= 0.72, 0.82 p=0.04, 0.01 respectively).
One ARMS subject in the P300, and one additional ARMS subject in the MMN group underwent transition to psychosis subsequent to taking part in the 1H-MRS and EEG studies.
Discussion
This is the first study to investigate the relationship between auditory evoked potentials and brain neurochemistry in individuals at high risk of psychosis. This was an exploratory analysis, and due to the small numbers included in this study, results should be viewed as preliminary and will require replication.Furthermore, the time between EEG and MRS acquisitions means that findings from this study represent trait, and not to state, relationships.
We found that lower thalamic Glx levels were related to frontal MMN amplitude, suggesting that deviance detection by the brain relies on glutamatergic mechanisms, as previously demonstrated with the NMDA receptor antagonist ketamine (Umbricht et al. 2000). Lower thalamic NAA levels were associated with more abnormal P300 response, as previously hypothesised(Molina et al. 2002), suggesting that loss of neuronal integrity in thalamus may underlie impairments in the separation of personally meaningful from irrelevant stimuli in schizophrenia. These findings are in agreement with our previous speculation that glutamatergic abnormalities might occur earlier in the illness, with changes in neuronal integrity occurring later, secondary to excitotoxic processes or other mechanisms (Stone et al. 2007).
Conclusions
Theseresults support a thalamic basis for filtering abnormalities in psychosis, as indexed by frontal P300 and MMN (Klostermann et al. 2006). Future work will determine whether MMN, P300 or thalamic spectroscopy measures, either individually or combined, might predict transition to psychosis.
Conflict of interest: none
References
Andreasen, N. C., 1999. A unitary model of schizophrenia: Bleuler's "fragmented phrene" as schizencephaly. Arch Gen Psychiatry 56, 781-787.
Bramon, E., Croft, R. J., McDonald, C., Virdi, G. K., Gruzelier, J. G., Baldeweg, T., Sham, P. C., Frangou, S. and Murray, R. M., 2004. Mismatch negativity in schizophrenia: a family study. Schizophr Res 67, 1-10.
Bramon, E., Shaikh, M., Broome, M., Lappin, J., Berge, D., Day, F., Woolley, J., Tabraham, P., Madre, M., Johns, L., Howes, O., Valmaggia, L., Perez, V., Sham, P., Murray, R. M. and McGuire, P., 2008. Abnormal P300 in people with high risk of developing psychosis. Neuroimage 41, 553-560.
Frangou, S., Sharma, T., Alarcon, G., Sigmudsson, T., Takei, N., Binnie, C. and Murray, R. M., 1997. The Maudsley Family Study, II: Endogenous event-related potentials in familial schizophrenia. Schizophr Res 23, 45-53.
Friedman, D., Cycowicz, Y. M. and Gaeta, H., 2001. The novelty P3: an event-related brain potential (ERP) sign of the brain's evaluation of novelty. Neurosci Biobehav Rev 25, 355-373.
Klostermann, F., Wahl, M., Marzinzik, F., Schneider, G. H., Kupsch, A. and Curio, G., 2006. Mental chronometry of target detection: human thalamus leads cortex. Brain 129, 923-931.
Molina, V., Reig, S., Desco, M., Gispert, J. D., Sanz, J., Sarramea, F., Pascau, J., Benito, C., Martinez-Lazaro, R., Luque, R., Aragues, M., Misiego, J. M., Corral, I. L. and Palomo, T., 2002. Multimodal neuroimaging studies and neurodevelopment and neurodegeneration hypotheses of schizophrenia. Neurotox Res 4, 437-451.
Phillips, L. J., Yung, A. R. and McGorry, P. D., 2000. Identification of young people at risk of psychosis: validation of Personal Assessment and Crisis Evaluation Clinic intake criteria. Aust N Z J Psychiatry 34 Suppl, S164-169.
Provencher, S. W., 1993. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn Reson Med 30, 672-679.
Semlitsch, H. V., Anderer, P., Schuster, P. and Presslich, O., 1986. A solution for reliable and valid reduction of ocular artifacts, applied to the P300 ERP. Psychophysiology 23, 695-703.
Sharp, F. R., Tomitaka, M., Bernaudin, M. and Tomitaka, S., 2001. Psychosis: pathological activation of limbic thalamocortical circuits by psychomimetics and schizophrenia? Trends Neurosci 24, 330-334.
Stone, J. M., Day, F., Tsagaraki, H., Valli, I., McLean, M. A., Lythgoe, D. J., O'Gorman, R. L., Barker, G. J. and McGuire, P. K., 2009. Glutamate Dysfunction in People with Prodromal Symptoms of Psychosis: Relationship to Gray Matter Volume. Biol Psychiatry.
Stone, J. M., Morrison, P. D. and Pilowsky, L. S., 2007. Glutamate and dopamine dysregulation in schizophrenia--a synthesis and selective review. J Psychopharmacol 21, 440-452.
Sumich, A., Kumari, V., Dodd, P., Ettinger, U., Hughes, C., Zachariah, E. and Sharma, T., 2008. N100 and P300 amplitude to Go and No-Go variants of the auditory oddball in siblings discordant for schizophrenia. Schizophr Res 98, 265-277.
Umbricht, D., Schmid, L., Koller, R., Vollenweider, F. X., Hell, D. and Javitt, D. C., 2000. Ketamine-induced deficits in auditory and visual context-dependent processing in healthy volunteers: implications for models of cognitive deficits in schizophrenia. Arch Gen Psychiatry 57, 1139-1147.
Figure 1
Relationship between thalamic glutamate+glutamine (Glx) and mismatch negativity at F3 and FZ.
Figure 2
Relationship between thalamic N-acetylaspartate (NAA) and P300 at F4.