Transient N-Methyl-D-Aspartate Receptor Blockade in
Early Development Causes Lasting Cognitive Deficits
Relevant to Schizophrenia
Mark Renato Stefani and Bita Moghaddam
Background: Aberrant N-methyl-D-aspartate (NMDA) receptor-mediated glutamatergic transmission has been implicated in
schizophrenia. We studied whether transient inhibition of NMDA receptor activity during early postnatal development would produce
a behavioral phenotype resembling that of individuals who are susceptible to develop schizophrenia.
Methods: Rat pups were given injections of the NMDA channel blocker MK801 on postnatal days 7 through 10. This period is akin
to the prenatal second trimester of primate development. Cognitive function was tested in adulthood.
Results: Treatment with MK801 impaired cognitive flexibility and working memory. The impairment in cognitive flexibility was due
to increased perseverative behavior. Treatment did not affect locomotor activity or recognition memory.
Conclusions: These results suggest that a brief disruption of NMDA receptors during a sensitive period of cortical development is
sufficient to produce selective cognitive deficits that are relevant to schizophrenia.
Key Words: Glutamate, memory, MK801, prefrontal, set-shift, spon-
taneous alternation
C
onverging lines of evidence implicate aberrant N-methyl-
D-aspartate (NMDA) receptor function in schizophrenia
(Javitt and Zukin 1991; Tamminga 1998; Tsai and Coyle
2002). N-methyl-D-aspartate receptor antagonists produce
schizophrenialike symptoms and cognitive deficits in healthy
humans (Adler et al 1999; Krystal et al 1994). Many of the genes
that are linked to the susceptibility to develop schizophrenia are
associated with proteins that modulate the function of NMDA
receptors (Harrison and Owen 2003; Moghaddam 2003).
N-methyl-D-aspartate receptors play a critical role in brain
development (Ritter et al 2002; Sircar 2000). Consistent with the
notion of schizophrenia being a developmental disorder (Wein-
berger 1987), blockade of NMDA receptors during early devel-
opment produces structural and behavioral abnormalities that
are relevant to schizophrenia (Harris et al 2003; Ikonomidou et al
1999; Luthi et al 2001; Sircar 2003).
We sought to determine whether a brief disruption of NMDA
receptor function during a critical stage of development, akin to
that which might be caused by transient epigenetic factors, is
sufficient to produce cognitive deficits like those associated with
schizophrenia. N-methyl-D-aspartate receptor channels were
blocked during a postnatal period critical to the development of
circuits involving the frontal cortex. Four measures of cognition
relevant to the behavioral traits of individuals susceptible to
develop schizophrenia were evaluated once rats reached adult-
hood.
Methods and Materials
Subjects and Treatment Schedule
Male Sprague-Dawley rats (n = 44) from 14 litters born to
different dams (Harlan, Somerville, New Jersey) were used. On
postnatal day 6 (P6), pups were randomly assigned to one of two
treatment groups: vehicle (saline, 1 mL/kg) or MK801 (.1 mg/kg
per injection). The pups were weighed daily from P6 through
P11 and weekly thereafter.
Beginning on P7, pups received subcutaneous injections
twice daily (09:00 and 16:00) for 4 days. Pups were removed
from their dam for 1 hour for the injection, during which time
they were housed in a tray placed on a heating pad maintained
at 37°C. The experimenters remained blind to the treatment until
the end of the study.
Pups were weaned at P20 and housed with their litter mates,
three or four per cage. Beginning on P60, they were tested for
locomotor and exploratory activity, spatial and nonspatial mem-
ory, and behavioral flexibility. All rats were tested in the same
sequence, although not all rats were tested on all tasks. Animal
use procedures were in accordance with the NIH Guide for the
Care and Use of Laboratory Animals.
Open Field Procedure. Rats were allowed 10 minutes of
exploration in an open field box (1 m x 1 m, divided into 25
squares). An experimenter rated numbers of square entries,
rears, forays into the open field, and time spent in the open field.
Object Recognition Procedure. The object recognition test
consisted of two phases, a study phase (exposure to two
identical objects) and a test phase (exposure to an object
identical to the first two objects and a novel object), separated by
a 2-hour retention interval. An object recognition score (N/T
score) was calculated as the ratio of the time spent exploring the
novel object (N) to the total time spent exploring both objects
(T). The locations of the familiar and novel objects were coun-
terbalanced within the open field chamber between rats.
Spontaneous Alternation Procedure. Rats were allowed 15
minutes of unimpeded exploration in a four-arm maze. The
number and sequence of arm entries was recorded for calcula-
tion of a percent alternation score (Stefani and Gold 2001).
Cognitive Set-Shifting Procedure. Set-shift testing was con-
ducted in a four-arm maze, the arms of which varied by texture
and brightness (Stefani et al 2003). The test consisted of two
sessions or sets. On set 1, rats were trained on a brightness
discrimination task (dark vs. light maze arms) to a criterion
performance level of eight consecutive correct entries. The
following day, the rats were trained in the same apparatus on set
2, a texture discrimination task (rough vs. smooth arms) for 80
From the Department of Neuroscience, University of Pittsburgh, Pittsburgh,
Pennsylvania.
Address reprint requests to Mark R. Stefani, Ph.D., Department of Neuro-
science, University of Pittsburgh, 446 Crawford Hall, Pittsburgh, PA
15260; E-mail: stefani@bns.pitt.edu.
Received January 16, 2004; revised May 6, 2004; revised November 15, 2004;
accepted November 18, 2004.
BIOL PSYCHIATRY 2005;57:433– 436 0006-3223/05/$30.00
doi:10.1016/j.biopsych.2004.11.031 © 2005 Society of Biological Psychiatry