Nonhuman primate model of schizophrenia using
a noninvasive EEG method
Ricardo Gil-da-Costa1, Gene R. Stoner, Raynard Fung, and Thomas D. Albright1
Systems Neurobiology Laboratories, Salk Institute for Biological Studies, La Jolla, CA 92037
Contributed by Thomas D. Albright, July 5, 2013 (sent for review March 26, 2013)
There is growing evidence that impaired sensory-processing sig-
nificantly contributes to the cognitive deficits found in schizo-
phrenia. For example, the mismatch negativity (MMN) and P3a
event-related potentials (ERPs), neurophysiological indices of
sensory and cognitive function, are reduced in schizophrenia
patients and may be used as biomarkers of the disease. In agree-
ment with glutamatergic theories of schizophrenia, NMDA antag-
onists, such as ketamine, elicit many symptoms of schizophrenia
when administered to normal subjects, including reductions in the
MMN and the P3a. We sought to develop a nonhuman primate
(NHP) model of schizophrenia based on NMDA-receptor blockade
using subanesthetic administration of ketamine. This provided
neurophysiological measures of sensory and cognitive function
that were directly comparable to those recorded from humans.
We first developed methods that allowed recording of ERPs from
humans and rhesus macaques and found homologous MMN and
P3a ERPs during an auditory oddball paradigm. We then investi-
gated the effect of ketamine on these ERPs in macaques. As found
in humans with schizophrenia, as well as in normal subjects given
ketamine, we observed a significant decrease in amplitude of both
ERPs. Our findings suggest the potential of a pharmacologically
induced model of schizophrenia in NHPs that can pave the way for
EEG-guided investigations into cellular mechanisms and therapies.
Furthermore, given the established link between these ERPs, the glu-
tamatergic system, and deficits in other neuropsychiatric disorders,
our model can be used to investigate a wide range of pathologies.
from neuronal pathology in multiple brain systems (1). Current
theories suggest that some of the sensory and cognitive symp-
toms of schizophrenia may, at least partially, result from dysfunc-
tion of the glutamate neurotransmitter system (2). In support of
this theory, it has been found that acute subanesthetic doses of the
N-methyl-D-aspartate receptor (NMDAR) antagonist ketamine
induces sensory and cognitive deficits akin to those experienced
by schizophrenia patients, as well as decreases of the mismatch
negativity (MMN) and P3 event-related potential (ERP) ampli-
The MMN is thought to reflect preattentive detection of
a deviant stimulus (4), whereas the P3 is thought to reflect the
redirection of attention to that deviant stimulus (5). In an odd-
ball paradigm, responses to deviant (or “oddball”) stimuli oc-
curring among a sequence of standard stimuli are measured. The
MMN is obtained by subtracting the ERP to the standard stim-
ulus from the ERP to the deviant stimulus, whereas the P3a is
typically observed in the ERP to deviants.
Schizophrenia patients appear less able to detect and direct
attention to novel stimuli than healthy controls (6). Consistent
with this behavioral deficit, the amplitudes of both the MMN (7)
and the P3 (8) have been found to be reduced in schizophrenia
patients, leading to the proposals that reduced MMN is a marker
of progressive pathology (7) and that reductions in both MMN
and P3a are markers of vulnerability for this disorder (8, 9).
Given the homology of human and rhesus macaque brains
(10), the development of a nonhuman primate (NHP) model of
chizophrenia is a multifaceted disorder that may originate
schizophrenia holds great potential for understanding the un-
derlying cellular pathophysiologies and for exploring potential
treatments. Of particular importance is the development of meth-
ods that allow comparison of neurophysiological correlates of
sensory and cognitive functions in NHPs and humans. To this
end, we developed a noninvasive electroencephalography (EEG)
system that uses common recording hardware and analyses for
the two species. Our system uses a noninvasive EEG cap in
NHPs, with electrode density identical to that used in humans.
Our approach allows for the calculation of topographic voltage
maps and localization of activity generators in the NHP brain.
To determine the utility of our NHP EEG system, we recorded
ERPs from humans (64-electrode array) (Fig. S1A) and NHPs
(22-electrode array) (Fig. S1B) during a passive auditory intensity
oddball paradigm. For both species, we established that ERPs
had timing and topographic distributions consistent with pre-
vious reports, and source localization suggested homologous
neural generators. Next, we investigated the effect of transient
administration of subanesthetic doses of ketamine on these
components in NHPs. These experiments revealed transient
but selective reductions of MMN and P3a components, which
mimicked those previously seen in human subjects similarly
treated with NMDAR blockers. Most significantly, they also
mimicked the chronic MMN and P3a reductions characteristic
Our findings, thus, support the utility of this NHP EEG sys-
tem, used in conjunction with a ketamine-administration model
of schizophrenia, to assay sensory and cognitive deficits. Our
approach can, thus, be used to facilitate understanding of neural
circuit dysfunctions characteristic of schizophrenia. Additionally,
a wealth of previous evidence has shown a significant correlation
between behavioral deficits and modulations of the MMN and
P3a ERPs in a variety of neurological and neuropsychiatric pa-
thologies (e.g., Alzheimer’s disease, dementia, Parkinson dis-
ease, affective disorders, and disorders of consciousness, etc.) (7,
11–13). Thus, our approach may also enable exploration, at
neuronal and behavioral levels, of therapies targeted at this
collection of pathologies.
Comparison of MMN in Humans and Monkeys. The MMN is
obtained by subtracting the ERP to the standard stimulus from
the ERP to the deviant stimulus (see Materials and Methods). In
humans, the auditory MMN is well documented as a fronto-
central negative potential with a latency of 100–250 ms after
the onset of stimulus presentation and has sources in auditory
cortices and in the inferior frontal gyrus (14). Consistent with
Author contributions: R.G.-d.-C., G.R.S., R.F., and T.D.A. designed research; R.G.-d.-C. and
R.F. performed research; R.G.-d.-C. and R.F. analyzed data; and R.G.-d.-C., G.R.S., R.F., and
T.D.A. wrote the paper.
The authors declare no conflict of interest.
Freely available online through the PNAS open access option.
See Commentary on page 15175.
1To whom correspondence may be addressed. E-mail: email@example.com or firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| September 17, 2013
| vol. 110
| no. 38
were trained to maintain central fixation. The fixation target was a red circle
(1° in diameter) on a black background presented using a 21-inch Sony GDM-
C520 CRT monitor at a 40-cm viewing distance.
EEG Data Collection/Recordings. For both human and NHP subjects, EEG scalp
recordings were acquired with the Vision Recorder software (Brain Products)
using a BrainAmp MR amplifier (Brain Products). We used a 64-channel EEG
cap BrainCap MR (Brain Products) with Ag/AgCl electrodes for human subject
data collection and customized 22-channel EEG caps, also with Ag/AgCl
electrodes, for NHPs. Collection of NHP EEG data required several additional
steps (SI Materials and Methods). NHPs were restrained in the chair in
a sphinx-like position with head protruding, stabilized, and facing forward.
EEG Data Analysis. EEG data were analyzed using Analyzer 2.0 software (Brain
Products). The analysis procedure included preprocessing (rereferencing the
datasets, band-pass filtering, down-sampling, segmentation, etc.) before
calculating ERPs for each condition. The same analyses were applied for
humans and NHPs.
Identification of Human and NHP ERPs. We first identified MMN and P3a
components in humans and then searched for homologous components in
NHPs before pharmacological manipulation. ERP components were identified
using established criteria. MMN was defined as the difference wave obtained
by subtracting ERPs for standard from ERPs for deviant stimuli. The P3a was
identified and analyzed from deviant stimulus trials. We ascertained the
timing, electrode location, voltage scalp distribution, and neural generators
for these ERP components. A 40-ms time window was placed around the
maximal amplitude in the average ERP waveforms of each species and was
used to extract mean amplitude values per subject from single trials. These
values were used for statistical analysis [MMN, two-way repeated-measures
ANOVA (factor 1, standard vs. deviant; factor 2, high vs. low); P3a, t test of
response to deviants] (STATISTICA data analysis software, 2007; StatSoft).
Ketamine and Saline Injections. Using the same passive auditory intensity
oddball paradigm EEG data were collected from two NHPs under three
physiological conditions: (i) “ketamine” (injection of ketamine; 1 mg/kg); (ii)
“saline” (injection of saline solution); and (iii) “5 h postketamine” (injection
of ketamine; 1 mg/kg). All injections were i.m. Recording began 12 min after
injection for ketamine and saline conditions and 5 h after injection for 5 h
postketamine condition. All recording sessions lasted 18 min. NHPs showed
no behavioral signs of ketamine effects (i.e., no signs of drowsiness and no
differential behavior between ketamine and saline conditions). A 40-ms time
window was established around the maximal amplitude in the average ERP
(MMN and P3a) waveforms and was used to extract mean amplitude values
per subject from single trials. These values were used for statistical analysis
[MMN, three-way repeated-measures ANOVA (factor 1, physiological con-
dition; factor 2, standard vs. deviant; factor 3, high vs. low tone); P3a two-
way repeated-measures ANOVA (factor 1, physiological conditions; factor 2,
high vs. low)] (STATISTICA data analysis software, 2007; StatSoft).
Topographic Voltage Maps and Source Analysis. Topographic voltage-distri-
bution maps for both human and NHP data were calculated in Cartool 3.43
(D. Brunet, Functional Brain Mapping Laboratory, Geneva, Switzerland) using
previously acquired electrode-position files for the 64-channel human and
22-channel NHP caps. Estimation of intracranial generators for MMN and P3a
was performed using Cartool 3.43 software with LORETA. Neural generators
were estimated across two time intervals per species: human (56–188 ms and
208–256 ms) and NHP (48–120 ms and 104–248 ms) corresponding to the
MMN and P3a components, respectively.
ACKNOWLEDGMENTS. We thank Steven Hillyard, Antigona Martinez, and
Marla Zinni for valuable contributions to design and data analysis; Thomas
Liu and Valur Olafsson for assistance in EEG setup; and Dinh Diep and Aaron
Cortez for assistance in animal training and care. Additionally, we thank
Denis Brunet for assistance with developing NHP inverse solutions. Stimulus
presentation for this experiment was conducted using the Cogent 2000
and Cogent graphics software (MATLAB toolbox), developed by teams at
the Wellcome Department of Imaging Neuroscience and University College
London. Cartool software (http://brainmapping.unige.ch/cartool) was pro-
grammed by Denis Brunet (Functional Brain Mapping Laboratory) and sup-
ported by the Center for Biomedical Imaging of Geneva and Lausanne.
1. Rissling AJ, Light GA (2010) Neurophysiological measures of sensory registration,
stimulus discrimination, and selection in schizophrenia patients. Curr Top Behav
2. Javitt DC, Zukin SR (1991) Recent advances in the phencyclidine model of schizo-
phrenia. Am J Psychiatry 148(10):1301–1308.
3. Umbricht D, et al. (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(12):1139–1147.
4. Garrido MI, Kilner JM, Kiebel SJ, Friston KJ (2007) Evoked brain responses are gen-
erated by feedback loops. Proc Natl Acad Sci USA 104(52):20961–20966.
5. Sutton S, Braren M, Zubin J, John ER (1965) Evoked-potential correlates of stimulus
uncertainty. Science 150(3700):1187–1188.
6. Baribeau-Braun J, Picton TW, Gosselin JY (1983) Schizophrenia: A neurophysiological
evaluation of abnormal information processing. Science 219(4586):874–876.
7. Wynn JK, Sugar C, Horan WP, Kern R, Green MF (2010) Mismatch negativity, social
cognition, and functioning in schizophrenia patients. Biol Psychiatry 67(10):940–947.
8. van der Stelt O, Belger A (2007) Application of electroencephalography to the study
of cognitive and brain functions in schizophrenia. Schizophr Bull 33(4):955–970.
9. Jahshan C, et al. (2012) Cross-diagnostic comparison of duration mismatch negativity
and P3a in bipolar disorder and schizophrenia. Bipolar Disord 14(3):239–248.
10. Sereno MI, Tootell RBH (2005) From monkeys to humans: What do we now know
about brain homologies? Curr Opin Neurobiol 15(2):135–144.
11. Näätänen R, et al. (2011) The mismatch negativity: An index of cognitive decline in
neuropsychiatric and neurological diseases and in ageing. Brain 134(Pt 12):3435–3453.
12. Pekkonen E, Jousmäki V, Reinikainen K, Partanen J (1995) Automatic auditory dis-
crimination is impaired in Parkinson’s disease. Electroencephalogr Clin Neurophysiol
13. Vecchio F, Määttä S (2011) The use of auditory event-related potentials in Alzheimer’s
disease diagnosis. Int J Alzheimers Dis 2011:653173.
14. Garrido MI, Kilner JM, Stephan KE, Friston KJ (2009) The mismatch negativity: A re-
view of underlying mechanisms. Clin Neurophysiol 120(3):453–463.
15. Javitt DC, Schroeder CE, Steinschneider M, Arezzo JC, Vaughan HG, Jr. (1992) Dem-
onstration of mismatch negativity in the monkey. Electroencephalogr Clin Neuro-
16. Woodman GF (2011) The Oxford Handbook of Event-Related Potential Components,
eds Kappenman ES, Luck SJ (Oxford University Press, New York).
17. Polich J (2007) Updating P300: An integrative theory of P3a and P3b. Clin Neuro-
18. Wieber J, Gugler R, Hengstmann JH, Dengler HJ (1975) Pharmacokinetics of ketamine
in man. Anaesthesist 24(6):260–263.
19. Stoet G, Snyder LH (2006) Effects of the NMDA antagonist ketamine on task-switching
performance: Evidence for specific impairments of executive control. Neuropsycho-
20. Taffe MA, Davis SA, Gutierrez T, Gold LH (2002) Ketamine impairs multiple cognitive
domains in rhesus monkeys. Drug Alcohol Depend 68(2):175–187.
21. Meltzer HY, Stahl SM (1976) The dopamine hypothesis of schizophrenia: A review.
Schizophr Bull 2(1):19–76.
22. Goldberg TE, Weinberger DR (1996) Effects of neuroleptic medications on the cog-
nition of patients with schizophrenia: A review of recent studies. J Clin Psychiatry
23. Foussias G, Remington G (2010) Antipsychotics and schizophrenia: From efficacy and
effectiveness to clinical decision-making. Can J Psychiatry 55(3):117–125.
24. Korostenskaja M, Kähkönen S (2009) What do ERPs and ERFs reveal about the effect
of antipsychotic treatment on cognition in schizophrenia? Curr Pharm Des 15(22):
25. Luby ED, Gottlieb JS, Cohen BD, Rosenbaum G, Domino EF (1962) Model psychoses
and schizophrenia. Am J Psychiatry 119:61–67.
26. Zukin SR, Zukin RS (1979) Specific [3H]phencyclidine binding in rat central nervous
system. Proc Natl Acad Sci USA 76(10):5372–5376.
27. Kantrowitz JT, Javitt DC (2010) N-methyl-d-aspartate (NMDA) receptor dysfunction or
dysregulation: The final common pathway on the road to schizophrenia? Brain Res
28. Swerdlow NR (2011) Are we studying and treating schizophrenia correctly? Schizophr
29. Seeman P, Guan H-C (2008) Phencyclidine and glutamate agonist LY379268 stimulate
dopamine D2High receptors: D2 basis for schizophrenia. Synapse 62(11):819–828.
30. Javitt DC, Steinschneider M, Schroeder CE, Arezzo JC (1996) Role of cortical N-methyl-
D-aspartate receptors in auditory sensory memory and mismatch negativity genera-
tion: Implications for schizophrenia. Proc Natl Acad Sci USA 93(21):11962–11967.
31. Umbricht D, Koller R, Vollenweider FX, Schmid L (2002) Mismatch negativity predicts
psychotic experiences induced by NMDA receptor antagonist in healthy volunteers.
Biol Psychiatry 51(5):400–406.
32. Kähkönen S, et al. (2002) Dopamine modulates involuntary attention shifting and
reorienting: An electromagnetic study. Clin Neurophysiol 113(12):1894–1902.
33. Ueno A, et al. (2008) Auditory ERPs to stimulus deviance in an awake chimpanzee
(Pan troglodytes): Towards hominid cognitive neurosciences. PLoS ONE 3(1):e1442.
34. Wylie GR, Clark EA, Butler PD, Javitt DC (2010) Schizophrenia patients show task
switching deficits consistent with N-methyl-d-aspartate system dysfunction but not
global executive deficits: Implications for pathophysiology of executive dysfunction in
schizophrenia. Schizophr Bull 36(3):585–594.
35. Takahashi H, et al. (2012) Neural substrates of normal and impaired preattentive
sensory discrimination in large cohorts of nonpsychiatric subjects and schizophrenia
patients as indexed by MMN and P3a change detection responses. Neuroimage 66C:
36. Wacongne C, Changeux J-P, Dehaene S (2012) A neuronal model of predictive coding
accounting for the mismatch negativity. J Neurosci 32(11):3665–3678.
| www.pnas.org/cgi/doi/10.1073/pnas.1312264110Gil-da-Costa et al.