Authors: Perminder S. Sachdev

Secondary Schizophrenia (61 page)

patients with schizophrenia. In a landmark study,
Carbamazapine, which does not appreciably reduce
acute ill patients were randomized to LY2140023 (an
glutamate release despite its similar action to lamo-orally absorbable analogue of LY354740), olanzapine,
trigine in blocking sodium channels
[263],
is ineffec-or placebo. This study demonstrated that LY2140023

tive as an augmenting agent in treatment-resistant
was effective against the positive and negative symp-schizophrenia.

toms of schizophrenia and had few side effects
[273].

Of greater interest is the predictive validity of
Hence, the PCP model showed accurate predictive
the PCP animal model in relation to new drugs
validity in the case of the first effective novel antipsy-that act directly on glutamate. Based on the PCP

chotic since the introduction of chlorpromazine.

model, researchers have now identified and tested
a range of new and promising compounds. Among
The PCP model of schizophrenia:

these is sarcosine, a glycine transport 1 (Glyt-1)
inhibitor, which increases glycine at the NMDA recep-

an integration

tor complex, thereby facilitating NMDA transmission.

When the psychotomimetic effect of PCP was first
Preclinical testing on PCP-induced models showed
proposed as a model of schizophrenia, the pharma-reversal of PCP effects
[264],
not as apparent in
cological actions of PCP were unknown. Although
the amphetamine-induced model. Subsequent clinian early report hypothesized impaired function of
cal testing revealed that sarcosine was ineffective as
glutamatergic neurons as a model of schizophrenia
monotherapy
[265],
but that it shows a significant ben-

[274],
it was not until the PCP binding site was local-eficial effect when used as an augmenting agent with
ized to the NMDA receptor complex that the NMDA
conventional antipsychotic treatment
[266].
Sarcosine
receptor hypofunction could be incorporated into the
did not offer additional benefit as an adjunct treat-PCP model
[253, 254].
A “thalamic filter dysfunc-ment with clozapine
[267]
, suggesting that clozapine
tion” was proposed
[253]
and pathological activa-may have direct glutamatergic actions. Another exam-tion of the cortico-striato-thalamocortical feedback
ple of drug development based on the PCP model
loop was hypothesized to cause information overload
is the preclinical testing of N-acetylaspartylglutamate
in the cortex (reviewed in [275]). Strengthening evi-

(NAAG) peptidase inhibitors. These compounds are
dence that the psychotomimetic effects of PCP were
selective group II mGluR agonists, which inhibit
directly related to NMDA receptor complex antag-presynaptic glutamate release
[268].
Preclinical testing
onism challenged the highly influential dopamine
of the NAAG peptidase inhibitor, ZJ43, showed that it
hypothesis of schizophrenia
[56].
An important ele-reduced MK-801-induced hyperlocomotion and PCP-ment of the PCP model, glutamatergic neuronal dis-induced stereotypic movements
[269],
effects due to
inhibition due to functional antagonism of NMDA

155

the mGluR3 agonist action of ZJ43
[270].
The field
receptors on GABAergic interneurons that normally
Organic Syndromes of Schizophrenia – Section 3

Figure 10.1
Panel A: Site of action of
PCP. G: GABAergic interneuron; P:

pyramidal neuron. Panel B: Circuits of the

brain relevant to the PCP model of

PFC

PCP

schizozphrenia. Black unbroken lines

+ + +

indicate glutamatergic

neurotransmission, the strength of which

AMG

is indicated by the number of
+
signs.

Black dotted lines indicate dopaminergic

neurotransmission. Grey unbroken lines

indicate serotonergic neurotransmission.

PFC: prefrontal cortex; AMG: amygdala; V.

+ + +

STRIATUM: ventral striatum; VP: ventral

STRIATUM

pallidum; THAL: thalamus; DMn:

dorsomedial nucleus of thalamus; An:

anterior nucleus of thalamus; HC:

THAL

hippocampus; MR: median raphe.

* = blocked inhibitory

DMn An

feedback to

pyramidal neurons

= NMDA receptor

complex

place excitatory pyramidal neurons under inhibitory
nucleus, and hippocampal-hippocampal-limbic stim-control, was added more recently
[276].
The puta-ulation of the anterior nucleus of the thalamus, protive role of cortical pruning
[277]
and evidence
vides an explanation for excessive subcortical-cortical
of reduced prefrontal neuropil
[278]
in schizophre-glutamatergic feedback drive to the prefrontal cortex
nia have also been incorporated into the model
(Figure 10.1,
Panel B).

[252].

The acute model has been supplemented by a

A number of neuropharmacological descriptions
chronic PCP model, which includes reduced pre-of the PCP model (hypoglutamatergic model) of
frontal extracellular glutamate and dopamine
[249].
A
schizophrenia have been published
[224,
279, 280, 281,

number of cellular mechanisms have been advanced to
282, 283].
Central to these models is a PCP-induced
link these surface receptor-focused models to intracel-deficit of GABAergic interneurons, which results
lular final common pathway models. Svenningson and
in disinhibition of glutamatergic pyramidal cells
colleagues
[200]
have implicated a common signal-

(Figure 10.1,
Panel A). Although this disinhibition is
ing pathway in the mediation of the psychotomimetic
assumed to be widespread throughout grey matter,
effects of glutamatergic antagonists (such as PCP),
models emphasize its impact on prefrontal cortex in
serotonergic agonists (such as LSD), and dopaminer-accounting for psychotomimetic effects. Increased lev-gic agonists (such as amphetamine). In this pathway,
els of extracellular glutamate in acute PCP models
phosphorylation status of Dopamine-and an Adeno-are thought to result from local collateral feedback
sine 3’,5’monophosphate (cAMP)-Regulated Phospho-by disinhibited pyramidal neurons onto presynap-Protein of 32 kilodaltons (DARPP-32) regulates down-

tic terminals, and increased nonNMDA glutamater-stream effector proteins, glycogen synthesis kinase-gic efferent feedback from the thalamus, other sub-3 (GSK-3), cAMP response element-binding proteins
cortical centers, and the hippocampal formation
(CREP) and c-Fos, thereby influencing electrophys-

(Figure 10.1,
Panel B). Completing the PCP model
iological, transcriptional, and behavioral responses.

are descriptions of increased dopamine efflux in the
An alternative mechanism for linking the PCP model
prefrontal cortex and ventral pallidum, resulting from
to intracellular signaling pathways is via glutamate-greater cortical drive to striatal/limbic subcortical cen-mediated excitotoxicity, which has been found to
ters. Increased cortical drive to the median raphe
induce apoptotic loss of dendrites and synapses with-results in increased prefrontal serotonin concentra-out cell body death or gliosis (reviewed in [283]). Post-tions, which augment glutamate efflux via presy-mortem studies have reported elevated Bax:Bcl-2 ratio
naptic 5HT2A receptor activation. Importantly, ven-

(a marker of increased propensity for apoptosis) in

156

tral pallidal stimulation of the dorsomedial thalamic
the temporal cortex
[284]
, comparable to functional
Chapter 10 – Psychotomimetic effects of PCP, LSD, and Ecstasy

findings in cultured fibroblasts from patients with
usually do not incorporate a neurodevelopmental per-schizophrenia
[285].
Taken together, these proposals
spective
[286, 287]
. This implies that alternative mod-provide a rich source of hypotheses for testing in stud-els may be required to complement pharmacological
ies of the pathogenesis of schizophrenia.

models (reviewed in [192]), preferably etiologically
linked to the occurrence of schizophrenia and based

Summary and conclusion

on a specific genetic alteration, exemplified by mouse
models of velo-cardio-facial syndrome
[288].

It is concluded that although PCP, LSD, and MDMA
In a more general sense, this review reminds
have well-documented psychotomimetic effects, only
us of our field’s historic dependence on research
for PCP is there abundant evidence that it induced
using patients. Preclinical models, especially in ani-psychotic disorder beyond the acute symptoms of
mals, release research from the inevitable confound-intoxication. In fact, there is no clear evidence that
ing factors related to illness experience and treatment.

either LSD or MDMA induces psychotic disorder, let
Most importantly, animal models allow direct obser-alone schizophrenia, in individuals who did not have
vation of neuronal dysfunction that models the human
vulnerability to schizophrenia premorbidly. Although
pathophysiology. Without this opportunity, major
inducing quantitatively less intense psychotomimetic
improvements in the drug treatment of schizophre-effects, ketamine is considered a safe and valid model
nia will not be possible. This review also reminded us
of PCP psychosis and applicable to preclinical human
of our dependence on clinical observation and mea-studies. We also concluded that rodent and primate
surement in developing adequately valid animal mod-models induced by PCP and analogues have con-els, which rely on the clinical insights of well-trained
struct validity, showing homologous behavior, cogni-clinicians who are interested in the neurobiology of
tive deficits, alteration in regional brain activation,
psychiatric disorder. Communication from the clinic
and underlying neuronal dysfunction, to PCP-induced
to the preclinical behavioral laboratory will enable the
psychotomimetic effects in healthy volunteers and
refinement of established models and the creation of
patients with schizophrenia. Most importantly, animal
new ones
[179].
The absence of well-validated, objec-models demonstrated predictive validity at the level
tive, and reliable measures of psychopathology is a bar-of hypothesis generation about human pathophysiol-rier to the development of homologous measurement
ogy and efficacy of novel drug therapies. Indeed, it
in animal models. Insufficient validation of an animal
could be said that PCP-related animal models have
model can be only as good as the information avail-been instrumental in the discovery of the first novel
able in the relevant preclinical human research and the
class of antipsychotic drug treatments (i.e., mGluR2/3

clinical literature. Clinical studies need to be informed
agonists) without D2R antagonist action since the
by results from animal studies as much as the reverse
introduction of chlorpromazine. The relative mer-is true. More translational science is needed to relate
its of PCP-related and amphetamine-induced mod-animal findings to humans and vice versa
[179].

els of schizophrenia have been discussed elsewhere
[223].
We considered both models to have high validity when descriptions and measurement are carried
Note

out at the level of neuronal dysfunction. A notable
Supplementary information related to this chapter
limitation of neurochemical models is that they
may be found at www.qsrf.com.au.

157

Organic Syndromes of Schizophrenia – Section 3

References

10. Sioris L. J., Krenzelok E. P.

phencyclidine. Concepts, Comp

Phencyclidine intoxication –

and Config, 1975.
4
(3):223–5.

1. Maddox V. H. (1981). The

literature review. Am J Hosp

historical development of

21. Fauman M. A., Fauman B. (1978).

Pharm, 1978.
35
(11):1362–7.