Research

The following can be found on these pages:

Report of Research Activities for 1996- The Stanley Laboratory of the Johns Hopkins University School of Medicine.

< span>NAMI Research Institute: The Stanley Foundation Neuropathology Consortium

< span>NAMI Research Institute: Information for Applicants For Using Postmortem Brain Tissue From the Stanley Foundation Neuropathology Consortium

Publications Stanley Lab

THE STANLEY LABORATORY OF THE
JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE
REPORT OF RESEARCH ACTIVITIES-1996

INTRODUCTION

The scientific goals of the Stanley Laboratory are based on the premise that human neuropsychiatric diseases arise from the interaction of infectious agents, the immune response, and the genetic susceptibility of an individual. This approach assumes that complex human diseases represent multiple interactions between environmental and genetic constituents. The elucidation of the etiopathogenesis of such diseases thus require general approaches and specific techniques which go beyond traditional searches for “the virus” or “the gene” thought to “cause the disease”. We have therefore undertaken a broad search for molecular species which are associated with the disease processes. These approaches involve a minimum of preconceived notions regarding the nature of the extraneous or endogenous components of the disease process. The theoretic bases for this approach is outlined in Table 1 in the manuscript “Viruses as Etiologic Agents of Schizophrenia” (Adv. Biol. Psychiatry 18:1-12, 1997).

The research performed in the laboratory has been totally supported by the Stanley Foundation. In addition to providing financial support, the Stanley Foundation, through the Stanley Neuropathology Consortium, has provided postmortem brain and cisternal fluid samples which have been crucial to the performance of the experiments described below. The work of the laboratory has benefited greatly from suggestions of many members of the faculty and staff of the NIMH Neurosciences Center as well as faculty members of the Johns Hopkins University School of Medicine.

HUMAN NEUROPSYCHIATRIC DISEASES
WORKING HYPOTHESIS

*Serious neuropsychiatric diseases are the result of environmental insults occurring in genetically susceptible individuals
*Environmental factors include infection and the immune response occurring during the prenatal or postnatal periods or later in life
*Genetic factors may include the determinants of the response to infection and the regulation of cytokines and other immune mediators.
*Different neuropsychiatric diseases such as schizophrenia and bipolar disorder may represent different combinations of these environmental and genetic factors.

I. DETECTION OF VIRAL PARTICLES IN GLIAL CELLS INOCULATED WITH BRAIN TISSUE FROM INDIVIDUALS WITH SCHIZOPHRENIA AND BIPOLAR DISORDER

Viruses in human tissue have traditionally been detected by propagation in tissue culture systems. This propagation is generally manifested by alterations of cellular architecture (cytopathic effects) visible by light microscopy. Previous studies have generally not detected such effects in cells inoculated with tissues obtained from individuals with neuropsychiatric diseases. However, the recent availability of brain tissue obtained postmortem from cases and controls as part of the Stanley Neuropathology Consortium have provided us with the opportunity to use more sensitive techniques to search for viral infection. One approach we have taken is to inoculate brain tissue obtained from cases and controls into the A172 neuroblastoma cell line and to examine the cells by sensitive electron microscopic, immunological, and molecular techniques for evidence of viral replication.

The initial results of these studies are described in Abstract #1 and the accompanying tables and photomicrographs. Electron microscopic examination of the cells indicated that four out of five tissue culture samples inoculated with brain tissue obtained from individuals with schizophrenia or bipolar disorder displayed alterations in cellular architecture typical of viral infections. These cells also contained nucleocapsid-like material similar to that seen in cells infected with viruses of the paramyxovirus class. Two of the samples also contained particles with morphology similar to that of human paramyxovirus (Figure 1). The presence of an antigen similar to paramyxovirus was confirmed by enzyme immunoassays employing A172 cells inoculated with case, but not control, brain tissue. Furthermore, cisternal fluids obtained postmortem from individuals with schizophrenia and bipolar disorder as part of the Stanley Neurovirology Consortium displayed a significant elevated level of antibodies to this antigen as compared to comparable control individuals. It is also of note that the type of viral particles seen in these experiments matches one of the RNA species identified by the differential display technique described in Abstract #1. On the other hand, performance of the polymerase chain reaction with primers directed at know paramyxoviruses such as measles, mumps, and parainfluenza virus did not yield any amplification. It is thus likely that the viral particles represents a new human strain of paramyxovirus. Additional studies are currently underway directed at the immunological, biological, and molecular characterization of this agent and the further determination of its role in human disease.

FIGURE 1:

This is a thin section of A172, a human glioblastoma cell line, inoculated with brain tissue from an individual with bipolar disorder, showing a virus-like particle budding from the cell membrane. The particle appears to be enveloped, 100-200 nm in size and may be intra- or extracellular. Such particles were not found in any cells inoculated with brains from normal individuals.

Manuscript “Viruses as Etiologic Agents of Schizophrenia can be found under the Publications Page.

ABSTRACT #1(see 1996 abstracts on Symposium page)
Detection of Viral Particles in Glial Cells Inoculated with Brain Tissue From Individuals with Schizophrenia and Bipolar Disease
I Dé, LM Melson, L. Brando, R. Yolken, EF Torrey, The Stanley Neuropathology Consortium

Cell Culture Experiments
Goal

* The identification of replicating viral agents in the brains of individuals with schizophrenia and bipolar disease.

Cell Culture Experiments

* The identification of a virus from the brains of individuals with neuropsychiatric disorders will allow us to better understand how these diseases evolve and provide another example of how viruses interact with their host to cause a chronic disease process.
*The definition of a role for viruses in the etiopathogenesis of psychiatric diseases may lead to new modalities for the prevention and treatment of these disorders.

Tissue Culture Methodology

*Cell

A172 Human Glioblastoma Cells

*Inoculum

0.2micron filtered brain extract pelleted by
ultracentrifugation at 123,000 xg

* Sampling

Cells and supernatants of mock and infected cells
harvested at 5 days post inoculation

* Passage

Samples obtained at 5 days repassaged in A172 cells

*Assays

Light microscopy
Electron Microscopy(EM)
Specific RT-PCR
Immunoassay methods

Ongoing Studies

*Cocultivation studies to identify susceptible host cell

MRC5
primary Rhesus monkey kidney
MDCK
CV-1

*Characterization of S39 antigen

peptide sequencing
polyclonal antibodies
2D electrophoresis

*Immunohistochemistry

antisera measles virus
antisera mumps virus
human sera

Summary and Conclusions

*Virus-like particles produced in cells inoculated with brains from schizophrenia and bipolar disease
*No such particles produced from control tissue
*Virus-like particle

apparently enveloped
10-200 nm morphologically similar to paramyxovirus
produce nucleocapsid material
does not induce cytopathic effect
do not cause productive infection
RT-PCR negative with primers directed at known paramyxoviruses

Cell Culture Experiments

Brain	Dx	City	Sx at TOD	Meds at TOD
S2	Sc	San Diego	Acute	Off Weeks/months
S3	Bi	San Diego	Remission	Prob. On
S11	Sc	Seattle	?Prob. symptomatic	Off Weeks/months
S39	Bi	San Diego	Prob. Acute	Off 1 week
S51	Sc	Seattle	Remission	On
S78	Nor	Minneapolis	—	—
S67	Nor	Settle	—	—
S58	Nor	Seattle	—	—

RT-PCR of RNA from Glioblastoma Cells

Primers	S2, S3, S39, S51, S78, Mock
BVDV	0
Panlentivirus	0
Borna Disease Virus	0
Enterovirus	0
All parainfluenza virus	0
Parainfluenza 3	0
Universal retrovirus	0
Morbillivirus	0

Cell Culture Experiments

	DX	Passage 1	Passage 2
S2	SC	Particles nucleocapsid-like inclusions, swollen endoplasmic reticulum (ER)	Normal cell morphology
S3	Bi	Swollen ER, nucleocapsid-like inclusions, no particles	No obvious pathology
S11	Sc	Swollen ER, no particles	No obvious pathology
S39	Bi	Particles, nucleocapsid-like inclusions	Budding particles
S51	Sc	No particles,normal cell morphology	N.D.
S78	Norm	No particles, normal cell morphology	N.D.
S58	Norm	No particles, normal cell morphology	N.D.
S67	Norm	No particles, normal cell morphology	N.D.

II. Viral and Virus-Related RNA Transcripts Are Differentially Expressed in the Brains of Individuals with Schizophrenia and Bipolar Disorder

As described in Abstract #2 and the subsequent figures and tables, we have utilized the technique of differential display to identify RNA species in the brains of individuals with schizophrenia and bipolar disorder which are not present in normal controls. We have also employed samples from individuals with Huntington’s disease to serve as controls for individuals who receive neuroleptic medications. One advantage of this approach is that it can identify RNA species which are differentially expressed in cases and controls without regard to the origin of these RNA. The RNAs can be of viral, microbial, or host origin.

We have currently identified four species which are expressed in greater concentrations in the brains of individuals with schizophrenia and bipolar disorder compared to the control groups. One of these RNAs is a transcriptional activator which binds to the tax protein of the virus HTLV-1. Another RNA is highly related to paramyxoviruses of humans and other primates. The other two RNA which are differentially expressed in the brains of affected individuals are homologous to retroviral sequences. One of the retroviral sequences is similar to primate group D retroviruses while the other is homologous to Rous Sarcoma Virus. These retroviruses could be either of exogenous or endogenous origin, that is they could represent either infectious retroviruses or be viruses which are part of the human genome but which have become activated in the brains of affected individuals. It is of note that group D retroviral sequences have been shown to integrate into the HLA-D locus of humans and other primates. The activation of this RNA might thus have significant consequences concerning immune activation and the generation of autoimmunity. The role of these RNAs, as well as others identified by this method, is currently under investigation. The methods are described in detail in “Identification of Differentially Expressed RNA Transcripts in Neuropsychiatric Disorders’ (Biol Psychiatry 41:759-761, 1997).

ABSTRACT #2 (see 1996 abstracts on Symposium Page)
Viral and Virus Related RNA Transcripts Are Differentially Expressed in the Brains of Individuals with Schizophrenia and Bipolar Disorder
F Yee, NL Johnston, F Leister, S Li, CA Ross, EF Torrey, RH Yolken, The Stanley Neurovirology Consortium

DIFFERENTIAL RNA EXPRESSION
Sequences of the FRAP-PCR Clones

Number of Clones	Sequencing Results
37/93 (40%)	Genes with known function = 18 Genes with unknown function (ESTs) = 9
7/93 (8%)	Repetitive sequences
12/93 (13%)	No significant matches
37/93 (40%)	Ribosomal RNAs

COMPARISON OF BRAIN RNA FROM SCHIZOPHRENIA CASE AND CONTROL

Sequence —————————-NO. (%) Unique

MatchGenes with	Clones Case > Control	Control > Case
Known Function	11 (30%)	2 (5%)
Unknown Function	11 (30%)	3 (9%)
Novel Sequence	8 (21%)	3 (5%)

Sequences of the FRAP-PCR Candidate Clones

Clone	Match	Blast n %ntd homology	Blast x % aa homology	Smith- Waterman
X6	DB1	>95	>95	N/A
X54	Pol/type D Retrovirus	N.S.	62 –11 (p <10)	N/A
X17	Env/SV-5 Paramyxovirus	N.S.	N.S.	P<0.01
X102	Env/avian Reticulosis Virus	N.S.	N.S.	P<0.001

Differential RNA Expression
Sequence Homologies

*Pol (=X54)

• group D retrovirus family
• similar to SRV-2 and MMTV
• related to human endogenous retroviruses (HERV-K)

*DB1 (=X6)

• zinc finger protein
• binds interleukin (IL)-3 promoter
• maintains basal IL-3 levels
• interacts with HTLV-1 Tax protein to stimulate IL-3 transcription

*SV-5 (=X17)

• paramyxovirus
• related to measles, mumps and parainfluenza viruses
• persistent in CNS infection

*AVIRE (=X102)

• retrovirus related to Rous sarcoma virus
• cause of generalized systemic infection in fowl
• related to endogenous retroviruses in humans

Differential RNA Expression
Ongoing Studies

*Further characterization of candidate FRAP-PCR clones by obtaining full length transcripts
*Confirmation of differential expression of candidate clones
*Examination of other brain regions to determine if the clones of interest are differentially expressed

Differential RNA Expression
Conclusions

*The FRAP-PCR techniques can be used to identify differentially expressed RNAs
*These results indicate that viral and virus-associated RNAs may be expressed in the frontal cortex of some individuals with schizophrenia and other psychiatric disorders

Differential RNA Expression
Screening of Candidate FRAP-PCR Clones

*Initial screening by RT-PCR of 37 candidate clones on frontal cortical RNAs from individual with:

– schizophrenia (n=7)
– bipolar disorder (n=7)
– normal controls (n=6)

*RT-PCR screening of a larger sample set of frontal cortical RNAs from individuals with:

– schizophrenia (n=21)
– bipolar disorder (n=19)
– depression without psychosis (n=13)
– Huntington’s Disease (n=6)
– normal controls (n=9)

For copy of manuscript “Identification of Differentially Expressed RNA Transcripts in Neuropsychiatric Disorders” go to Publications.

III. GENOMIC DIFFERENTIAL DISPLAY ENRICHED FOR RETROVIRAL LTR SEQUENCES

We are also exploring the possible role of retroviruses in human neuropsychiatric diseases by the analysis of DNA extracted from the lymphocytes of twins discordant for schizophrenia and bipolar disorder. These samples were collected by Dr. Torrey as part of his study of monozygotic twins discordant for schizophrenia and bipolar disorder. As described in Abstract #3 and the accompanying tables, we have devised a method which allows for the specific amplification of retroviral sequences by the use of the tRNA primer binding site. Since all retroviruses use tRNA binding as an integral part of their replication strategy, this method has the potential of allowing for the identification of novel retroviruses even when the exact retroviral sequence is unknown. Initial studies indicate that affected and unaffected twins differ in terms of at least 1 retroviral sequence. The characterization of this sequence and the identification of additional ones is ongoing.

ABSTRACT #3 (See 1996 abstracts on Symposium page)
Genomic Differential Display Enriched for Retroviral LTR Sequences
R Viscidi, E Rubalcaba and R Yolken

Human Endogenous Retroviruses

* “retroelements” in human genome resembling infectious retroviruses in genomic organization
* Viral particles, mRNA and anti-viral antibodies associated with human tumors, autoimmune and neurological diseases
*Pathogenic mechanisms in animal models: transcription of cellular genes, direct toxic effect, influence on immune function, interaction with infectious retroviruses.

Detection and Screening Methods

*Great number and complexity of retrovirus-like sequences in human genome poses challenge for methods development
*Search strategies:

low-stringency hybridization with probes derived from animal viruses
PCR using degenerate primers homologous to conserved regions of reverse transcriptase or protease gene

PCR Strategy for Differential Display of Genomic LTR Sequences

* Genomic DNA cleaved with a restriction enzyme and ligated to an oligo adapter
* DNA amplified with 18-mer corresponding to 3’terminus of a human tRNA and primer complementary to adapter
*PCR products resolved on agarose gel and analyzed by Fluorimager
*Difference bands excised, cloned, sequenced, and analyzed for homology to LTRs by BLAST

Advantages of PBS Primer for Amplification of LTR Sequences

*Number of host tRNAs is limited
*tRNAs used by many endogenous and exogenous retroviruses are known
*tRNA genes not amplified because 3′-CCA of mature tRNA is added post-transcriptionally
*Retroposons (LINEs), present in high copy number in human genome, are not amplified

Materials and Methods

*Subjects: monozygotic twins discordant for schizophrenia from NIMH twin collection
*Materials: lymphocyte DNA cut with Bgl II or Ban HI,ligated to oligo adapter (gift of Y. Sun)
*Primers: 18-mer corresponding to tRNA_Pro, tRNA_LYs(UUU), tRNA_Lys(CUU), or tRNA_His and 20-mer complementary to adapter
*PCR: 30 cycles at 95ºC – 55ºC -72ºC using pwo and taq mixture and 0.2uM primers

LTR differential display of genomic DNA of discordant twins

tRNA Primer	Adapter Primer	Subject	Difference Band	No. Clones	LTR match By BLAST
Lys (UUU)	Bgl II Bam HI		None	NA	NA
Lys (CUU)	Bgl II Bam Hi		None	NA	NA
Pro	BgII Bam Hi	Normal	700 bp	19	3
His	Bam HI Bgl II	Schizo.	1200 bp	4	2

Conclusions

* Primers corresponding to 3′ end of host tRNA can be used in a PCR strategy to amplify sequences containing LTR elements
*Because entire class or Retroviruses is amplified, the PCR can be used for differential display
*The utility of the technique to identify retroviral sequences associated with schizophrenia is currently under investigation.

IV. COMPARISON OF cDNA LIBRARIES FROM THE BRAINS OF NORMAL AND SCHIZOPHRENIC INDIVIDUALS SHOW THAT AN UNUSUALLY HIGH NUMBER OF NOVEL SEQUENCES ARE PRESENT IN THE BRAIN OF THE INDIVIDUAL WITH SCHIZOPHRENIA

We are exploring the differential expression of RNAs by the use of subtractive hybridization methods. As depicted in Abstract #4 and the accompanying tables, we have identified a number of RNA species which are present in the brains of individuals with schizophrenia but not controls. Several of these RNAs appear to be transcriptional activators similar to the one identified by the differential display technique. Other differentially expressed RNAs are apparent in the amplification reactions and are in the process of being identified. In order to increase our ability to identify novel RNAs, we have also developed and characterized libraries from the brains of individuals with schizophrenia and bipolar disorder. Our initial analyses of these RNAs indicate that a large number of the RNA species have not been previously identified in the human genome project or EST projects. This finding indicates that the brains of individuals with schizophrenia and, perhaps, other human neuropsychiatric disorders, may contain many RNA species not readily identifiable in samples commonly used for the human genome or EST initiatives. The availability of cDNA libraries derived from brain regions of individuals with schizophrenia and bipolar disease may provide a valuable resource for the study of these disease processes.

Abstract #4 (see 1996 Abstracts on Symposium Page

Comparison of cDNA Libraries from the Brains of Normal and Schizophrenic Individuals Show that an Unusually High Number of Novel Sequences are Present in the Brain of the Schizophrenic Individual
Nancy Johnston, Robert Yolken, E. Fuller Torrey, The Stanley Neuropathology Consortium

Subtracted Probes from Schizophrenia Frontal Lobe mRNA

ABSTRACT:
mRNA from postmortem human brains has been successfully used in the generation of cDNA for libraries, RT-PCR, and an assortment of subtractive and differential display techniques. The later methods are well suited to exploring the differences between mentally ill and unaffected individuals. We have used the Clonetech PCR-Select subtraction technique to generate a number of leads in the discovery of novel messages from the brains of schizophrenic individuals.

THE BACKGROUND:
A individual with mental illness has many irregularities in their brain mRNA compared to a matched mentally normal control.

THE GOAL:
*Use subtraction to identify messages that are overexpressed or unique to a schizophrenic brain.
*Screen against cDNA from a large set of samples, of varying diagnoses, to determine if messages are disease specific or elevated in a disease specific manner.

RESULTS: ROUND 2
Screen N432, 433, 435, 436, 438 against set of 20 samples (7 Scz, 6 PB, 1 D, 4 NC, 2HD)
Only 435 showed diagnosis-related elevations in levels.

RESULTS: ROUND 3
Screen N435 against set of 80 samples (19 Scz, 17 BP, 14 D, 8 NC, 6MD)

RESULTS: ROUND 1

Candidate Clones:	Approximate Size:	Identity:
N431 N432 N433 N434 N435 N436 N438 N450	550 bp 600 bp 300 bp 650 bp 2000+ bp 700 bp 400 bp 600 bp	Multiple ESTs Single human EST 2 ESTs multiple ESTs Zn finger/leu zipper No match at nucleotide or protein level No match at nucleotide level, at protein level some homology to bacterial toxin and to hormone receptor messages ??

MATERIALS AND METHODS
Demographic Information

Brain #	S-11	S-37
Diagnosis Age Gender Mode of Death PMI Cerebellar pH Other Notes	Schizophrenic 34 yrs. Male Suicide:jump 40 hrs. 5.9 no detectable neuroleptics in system at postmortem evaluation. Suspect individual was off meds for many months prior to his death.	Normal 28 years male cardiac 47 hrs. 6 Sudden death, seemingly healthy individual

MATERIAL AND METHODS
Library Construction

*Stratagene l ZAPII vector
*Total RNA from frontal lobe, purified over oligo dT columns
*Reverse transcribed with a mix of oligo dT and random hexamers primers
*Primary Clones:

-S-11: 1.45 x 10⁶ pfu
-S-37: 1.56 x 10⁶ pfu

CONCLUSIONS:
*The subtraction/amplification method can be used to identify differences in postmortem human brain poly A+ RNA
*Messages that appear to be elevated in brains of mentally ill individuals can be identified
*The screening procedure outlined above is an efficient method to identify candidate messages that are markers of mental illness.

ONGOING STUDIES:
*Further sequencing of candidate bands
*Screening of clones to determine disease specificity

V. SERIAL ANALYSIS OF GENE EXPRESSION OF HUMAN BRAIN TISSUE

We are exploring the range of expression of brain RNAs by the use of the technique of serial analyses of gene expression (SAGE) as described in Abstract #5 and the accompanying tables. These data indicate that more than 40% of the RNAs expressed in adult human brain do not have significant levels of homology with known human, mammalian, vertebrate, invertebrate, or microbial sequences. This finding indicates that currently available DNA and RNA libraries may not contain many of the RNA species which will prove to be important in the pathogenesis of important human psychiatric diseases such as schizophrenia and bipolar disorder.

Abstract #5 (see 1996 Abstracts on Symposium Page)
Serial Analysis of Gene Expression of Human Brain Tissue
Y Sun, RH Yolken, The Stanley Neuropathology Consortium

SERIAL ANALYSIS OF GENE EXPRESSION
Materials

Postmortem brain tissue: Frontal cortex

*Schizophrenia
*Bipolar disorder
*Normal

Specific Aims

*Analysis of gene expression in schizophrenia, bipolar disorder and normal brains
*Identification of disease-associated markers

*SAGE is an efficient method to analyze large numbers of transcripts in human brains in a short period of time.
*SAGE has the potential to detect qualitative and quantitative differences in RNA transcription occurring in cases and controls
*SAGE may be useful for the identification of disease-related genetic or viral-induced transcripts in complex human diseases such as schizophrenia and bipolar disorder.

Ongoing Studies

1. Continue to analyze transcripts from brain
2. Identification of differences in the abundance of gene population between each disease entity
3. Test candidate gene markers in selected cases and controls

Method

SAGE
Victor E. Velculescu, Lin Zhang et al, Science 270, 484-487, 1995

SAGE
Analysis of Tags from 13 Clones of Concatemers in Normal Human Brain

	Tags	Identification	Identification
Occurrence		Known	Unknown
4	1	0	1
3	1	1	0
2	13	4	9
1	136	38	98
Total	151	43	108

VI. CORRELATION OF SCHIZOPHRENIA AND BIPOLAR DISORDER WITH CYTOKINE AND CYTOKINE RECEPTORS IN CISTERNAL FLUIDS AND POSTMORTEM BRAIN TISSUE

Several previous studies have documented the activation of the immune system in many individuals with serious neuropsychiatric diseases. If the immune system is involved in these diseases, it is highly likely that cytokines and cytokine receptors play an important role in the mediation of the immune response. Cytokines are particularly important in establishing a link between viral infection and disease since a cytokine response can persist long after the clearance of virus by the host immune response. Furthermore, cytokines can mediate autoimmune reactions similar to those postulated in the etiology of some cases of human psychiatric diseases.

Previous studies have examined cytokine levels in the serum and lumbar cerebral spinal fluids of individuals with schizophrenia and bipolar disorder. However, the samples collected as part of the Stanley Neuropathology Consortium provided us the opportunity to examine cisternal cerebrospinal fluids obtained from the base of the brain in cases and controls. These fluids may be more representative of immune reactions in the brain than samples collected from more distal locations in the lumbar space. As described in Abstract #6 and the accompanying figures, we measured a large number of cytokine and cytokine receptors in such samples. We found that although there are individual variations in cytokine measurements, none of the standard cytokines showed statistically significant differences among the diagnostic categories. On the other hand, the principal cytokine receptor for interleukin-2 (il-2a) was significantly elevated in the fluids of individuals with schizophrenia as compared to controls. There were also some elevations in the fluids from individuals with bipolar disorder including individuals who had never received neuroleptic medication. These elevations were also not found in matching serum samples, indicating that the generation of the il-2a receptor was specific for the central nervous system. The finding of elevations in interleukin-2 receptors in the absence of elevation of cytokines is of interest since such elevations have been found in other persistent brain infections including subacute sclerosing panencephalitis (SSPE) occurring after measles infection as well as in central nervous infections due to HTLV-1. The exact anatomic location and cell-type responsible for the generation of the interleukin-2 receptors is currently under investigation.

Abstract #6(see 1996 Abstracts on Symposium Page)
Correlation of Schizophrenia and Bipolar Disorder with Cytokine and Cytokine Receptors in Ventricular Fluids and Postmortem Brain Tissue
L Bobo, N Novak, N Johnston, F Leister, EF Torrey, R Yolken, The Stanley Neuropathology Consortium

Introduction

*Epidemiological evidence suggests a possible infectious or autoimmune stimulus for schizophrenia
*Specific immunomodulators, that is cytokines, may play a role in brain development, pathology, neurotransmission and behavior
*Previous studies have shown abnormalities in certain cytokine and receptor levels from serum or CSF samples, and depressed peripheral blood mononuclear cell mitogenic response
*These abnormalities, while generally not linked to medication status, correlate with disease severity
*Ventricular fluids from cases and controls have not been previously analyzed

Immediate Ongoing Studies

* Identify brain area and cellular localization of IL-2R positive cells
*Survey brain areas for other immune markers of cellular activation
*Compare disease severity in schizophrenia with the observed high/low clusters for IL2 receptor levels

Some Examples of Infectious Etiologies Where Intrathecal or Long Duration Immune Activation Has Been Reported

*Sub-acute sclerosing panencephalitis
*Herpes simplex encephalitis
*Tropical spastic paraparesis/HTLV-1 associated myopathy

Cytokines Tested in Ventricular Fluid

*There was no statistically significant difference between the levels of interleukin-2 receptor alpha and the following potential confounding factors:

– Total lifetime medication (Fluphenazine units)
– Total lifetime hospitalization
– Age
– Gender
– Brain pH
– Agonal state
– Total protein
– Endotoxin
– Visible blood

Patients and Methods

*Patient population consisted of individuals with pre-mortem diagnoses of schizophrenia, bipolar disorder, non-psychotic depression, and controls with either no psychiatric disease or with Huntington’s Disease
*Cytokines and receptors were measured in ventricular fluids using enzyme-linked immunosorbant assays (ELISA), and levels are expressed as pg/ml
*Ventricular fluids were also screened for presence of endotoxin, blood and total protein
*Cytokines were measured in serum samples using ELISA
*Brain homogenates from the frontal pole were standardized to 500 ug/ml of protein for ELISA, and to GAPdH message for reverse transcriptase-PCR (RT-PCR) assays
*Data analysis was by ANOVA, t-test and multivariate regression

Cytokines Tested in Ventricular Fluids

* Interleukin-1 beta
* Interleukin-2
* Interleukin-2 Receptor alpha
* Interleukin-6
* Interleukin-6 Receptor
* Interleukin-10
* Interferon gamma
* Tumor Neurosis Factor alpha

Conclusions

*Interleukin-2 Receptor alpha is significantly elevated in ventricular fluids obtained post-mortem from individuals with schizophrenia
*Inflammatory cytokines, such as interleukin-1, interleukin-2, interferon gamma, and tumor necrosis factor were not significantly increased in cases as compared to controls
*Increased levels of interleukin-2 receptor alpha were not associated with potential confounding variables such as medication, pre-mortem, or post-mortem factors
*Serum levels of interleukin-2 receptor alpha were not significantly elevated in cases
*Interleukin-2 receptor alpha mRNA and protein were not detected in the frontal pole tissue obtained from individuals with schizophrenia

VII. MEASUREMENT OF RNA FROM 89 POSTMORTEM HUMAN BRAINS; A MULTIVARIATE STATISTICAL ANALYSIS OF PRE- AND POST MORTEM EFFECTS ON THE YIELDS OF GAPdH AS MEASURED BY RT-PCR

The studies described above rely heavily on post-mortem brain tissue obtained from the Stanley Neuropathology Consortium. It is important to document that such tissue can give valid results in molecular assays such as the ones being used in the course of our research. We have thus performed detailed experiments on the quality and quantity of RNA which can be obtained from brains collected in this manner. As described in Abstract #7 and the accompanying materials, we have found that low-abundance messenger RNA can be amplified from >90% of the brains collected as part of the consortium. The levels of RNA did not vary by clinical diagnosis; the RNAs obtained from the individuals with schizophrenia and bipolar disorder was indistinguishable in quality and quantity from that obtained from unaffected controls. We found that the quantity of the brain RNA was most dependent upon the pH of the brain tissue on receipt in the laboratory. On the other hand, the post-mortem interval or the conditions of storage were not correlated with RNA levels. This finding indicates that pre-mortem factors, such as brain hypoxia and artificial ventilation, may be the most important determinants of brain RNA quantity. We also found that the amount of RNA which could be amplified from the brain tissues by methods such as PCR was determined to a great extent by the methods used for the transcription of RNA to the corresponding cDNA. For example, transcription using oligo-dT complementary to the polyadenylation locus at the 3′ untranslated region of the RNA was highly affected by brain pH while transcription using random hexamers was relatively unaffected by pre-mortem or post-mortem factors. This finding is probably a reflection of the fact that the brain RNA is degraded from 3′ untranslated region of the message. This finding also indicates that the success of post-mortem measurements of RNA may depend, to a great extent, on the method used for RNA transcription.

Abstract 7 (see 1996 Abstracts on Symposium Page)
Measurement of RNA from 89 Postmortem Human Brains: A Multivariate Statistical Analysis of Pre- and Post Mortem Effects on the Yields of GAPdH As Measured by RT-PCR
N Johnston, J Cervenak, R Yolken, EF Torrey and The Stanley Neuropathology Consortium

Causes for RNA Level Variations in Human Brain
Introduction

*Background

-RNA levels in human brain tissue vary significantly between individuals
-Human brain RNA levels are consistently lower than animal brain levels purified concurrently
-Human tissues are subject to many pre- and postmortem influences which could affect levels of overall RNA
-pH has been shown as an indication of RNA levels¹

*Goal

-To determine which pre- and postmortem factors most influence RNA levels in human brain tissue
-To investigate whether the use of different primers in the reverse transcription reaction show different sensitivities to pre- and postmortem factors and pH
-To determine if there are differences in RNA levels based on mental diagnosis or modes of death

¹Harrison PJ et al, 1995
Kingsbury A et al, 1995

Conclusions

*GAPdH, a moderately expressed RNA, can be amplified from most postmortem brains from the Stanley Neuropathology Consortium by 3 different methods of RNA transcription
*pH is the strongest correlate of GAPdH amplification and affects all 3 methods of transcription
*Transcription primed by oligo-dT was the method most sensitive to pH and to other conditions of collection and storage. This was probably due to the premortem and postmortem degradation of the 3′ end of the RNA molecule
*Transcription primed by random hexamers was the transcription method least affected by pre- or postmortem factors
*None of the methods of RNA transcription was independently affected by:

-Clinical Diagnosis
-Age
-Gender
-Mode of death

Methods

*Generation of Data:

-Collect demographic information on all subjects
-Measure pH on occipital and cerebellar tissues from 89 brains
-Purify total RNA from occipital lobe tissue from the same brains
-Reverse transcribe equal amounts of RNA with 3 different primers (oligo dT, random hexamers, GAPdH specific primer)
-Perform quantitative PCR for glyceraldehyde phosphate dehydrogenase (GAPdH)

*Analysis of Data:

-Univariate Analysis (Spearman Rank)

-GAPdH vs demographics and pH

– Multivariate Analysis

Univariate Analysis of GAPdH Message vs. Demographic Factors
Correlation and significance values

Factors	Oligo dT primed cDNA	GAPdH Specific Primed cDNA	Random Hexamer primer cDNA
Female	-0.069	-0.035	0.084
Suicide	-0.25	-0.17	-0.1
CO Poisoning	0.19	0.26	0.26
Hospitalization	-0.15	-0.09	-0.14
Age	0.085	0.07	0.15
Freezer Interval	-0.31**	-0.19	0.098
Refrigeration Interval	-0.15	-0.13	-0.066
PMI	-0.2	-0.16	-0.059
Rapidity = 1	0.24	0.27**	0.14
Rapidity = 3.5	-0.25	-0.18	-0.26
Occipital pH	0.35***	0.34***	0.25

**p<0.01 ***p<0.001

Causes for RNA Level Variations in Human Brain
Comparisons of Various Priming
Methods

Primer	Advantages	Disadvantages
Oligo dT	No transcription of ribosomal or degraded mRNA	Sensitive to degradation of poly A tail Will not prime viral RNAs which lack poly A
Specific primers	No reliance on poly A+ tail can be used with “one-step” RT-PCR procedures	Requires separate priming reaction for each target Cannot be used in discovery of unknown targets
Random hexamers	Can transcribe all RNA	Will transcribe ribosomal and fragmented RNA

Conclusions

*pH is a rough indicator or RNA quantity
*3′ ends of RNA are more sensitive to various handling factors
*oligo dT primed reactions are sensitive to freezer interval and pH
*oligo dT primed reactions have the highest correlations with pH
*Specific priming correlates only with pH levels
*No effect by PMI, diagnosis, age, sex

Implications

*Postmortem brain tissue can be used for characterization of RNA expression from individuals with schizophrenia and bipolar disorder
*Methods which rely on poly A+ mRNA or on oligo dT priming methods are likely to be more sensitive to pre- and postmortem factors
*The analysis of postmortem brain tissues may lead to the discovery of RNAs related to the diagnosis, characterization and treatment of human neuropsychiatric disease

Factors Used In Multivariate Analysis

1. Age
2. Mentally Ill vs. Healthy
3. Suicide vs. Non-suicide
4. CO Death vs. Other Death
5. Hospitalization vs. Not
6. Freezer Interval
7. PMI
8. Occipital pH

VIII. CHARACTERIZATION OF DISEASE PROCESSES AND PHARMACEUTICAL MECHANISMS THROUGH QUANTITATIVE HIGH-THROUGHPUT TWO DIMENSIONAL Electrophoresis is likely that many disease-associated RNAs exert an effect on brain function following translation into proteins. Some of these proteins are likely to be present in the brains of affected individuals in concentrations of molecular species that differ from those of controls. We are thus performing a detailed analysis of proteins from the postmortem brain tissues of cases and controls by the use of standardized 2 dimensional electrophoresis, in collaboration with Dr. N. Leigh Anderson of Large Scale Technologies (Abstract #8). Initial analyses indicate that virtually all of the brain samples contain a full complement of proteins analyzable by this method (Figure 3). In addition, there appears to be several proteins which are present in increased concentrations in individuals with schizophrenia and bipolar disease. The identification of these proteins is in progress.

Abstract #8 (See 1996 abstracts on Symposium Page)
Characterization of Disease Processes and Pharmaceutical Mechanisms Through Quantitative High-Throughput 2-D Electrophoresis
NL Anderson, R Esquer-Blasco, G Jorgensen, C Sims, J Taylor, J-P Hofmann and NG Anderson (Large Scale Biology Corporation) N Johnston, R Yolken

IX. SUMMARY AND CONCLUSIONS

Our studies provide direct evidence for the involvement of environmental factors in the etiology and pathogenesis of some cases of schizophrenia and bipolar disorder. Possible environmental factors identified so far include infectious agents, particularly viruses of the paramyxovirus class, inflammatory responses, autoimmune reactions or some combination of these factors. Molecular studies of gene expression in brain tissue also indicate that these diseases are associated with the upregulation of a range of host factors many of which are involved in transcription and gene regulation. These findings are most consistent with a model in which infections interacts with host genetic factors to produce the disease phenotype. As outlined in figure 2, we believe that this model is consistent with many other human diseases which represent such interaction. For example, it is likely that many cases of cancer, rheumatic disorders, vascular diseases, and diseases of other organs arise from a complex interaction of genetic, infectious, immune, and other environmental factors. The elucidation of the etiopathogenesis of human psychiatric diseases such as schizophrenia and bipolar disorder might result in the development of effective methods for the diagnosis and treatment of these diseases, which have an enormous impact on human health in the United States and many areas of the world. Furthermore, these studies might lead to an increased understanding of many other complex human diseases which do not fall into simple categories of etiology or pathogenesis.

Figure 2