Journal of Neuroimmunology 214 (2009) 118124

Contents lists available at ScienceDirect

Journal of Neuroimmunology
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n e u r o i m

Antineuronal antibodies in OCD: Comparisons in children with OCD-only,

OCD+chronic tics and OCD+PANDAS
Colin Gause a, Christina Morris a, Shilpa Vernekar a, Carlos Pardo-Villamizar a, Marco A. Grados b,
Harvey S. Singer a,
a
b

Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD, United States
Department of Psychiatry, Johns Hopkins University School of Medicine, Baltimore, MD, United States

a r t i c l e

i n f o

Article history:
Received 3 March 2009
Received in revised form 16 June 2009
Accepted 17 June 2009
Keywords:
Obsessivecompulsive disorder
Autoimmunity
PANDAS
Tics
Antineuronal antibody

a b s t r a c t
Autoimmunity associated with a streptococcal infection has been proposed as a pathogenic mechanism for
obsessivecompulsive disorder (OCD) in children. Antibrain antibody proles were compared in children
with OCD-only (n = 13; 14.1 3.1 years), OCD+PANDAS (n = 20; 11.3 1.5 years), OCD+Chronic Tic
Disorder (n = 23; 13.4 3.5 years), and controls (n = 29; 12.4 2.4 years) using ELISA (orbitofrontal (OFC)
and dorsolateral prefrontal cortex (DLPFC), caudate (CD), cingulate gyrus (CG)), immunoblotting (four
regions plus putative antigens), and immunohistochemistry. ELISA and immunohistochemistry showed no
differences among groups. Immunoblot showed that a greater percentage of individuals in the OCD+PANDAS
cohort had reactive bands at 27 kDa (CD, CG, DLPFC), 36 kDa (CD), and 100 kDa (CD, OFC) and increased
peak height at 67 kDa (all regions). Immunoblotting studies using the putative antigens (pyruvate kinase M1,
aldolase C, - and -enolase) did not differ among groups. ASO titers were similar in all groups and did not
correlate with immunoassays. It remains controversial whether childhood OCD is associated with
autoimmune mechanisms.
2009 Elsevier B.V. All rights reserved.

1. Introduction
Obsessivecompulsive disorder (OCD) in children can occur in
isolation, or, as in the majority of cases, in association with other
disorders such as tics, Tourette syndrome (TS), Sydenham's chorea,
and pediatric autoimmune neuropsychiatric disorders associated with
streptococcal infections (PANDAS) (de Mathis et al., 2008). In part,
due to the diverse nature of this group, knowledge about the
pathogenesis of obsessivecompulsive symptoms in children remains
incomplete. An autoimmune etiology for OCD has been suggested
based primarily on three ndings: reports of Group A beta-hemolytic
streptococcal (GABHS) infections causing symptom exacerbations
(Mell et al., 2005; Murphy et al., 2004; Swedo et al., 1998); a high rate
of OCD in children with Sydenham's chorea (Mercadante et al., 2000;
Swedo, 1994; Swedo et al., 1989), PANDAS (Murphy and Pichichero,
2002; Swedo et al., 1998), and rheumatic fever without chorea
(Murphy et al., 2006); and the identication of putative pathogenic
antibrain antibodies (Church et al., 2002, 2004; Dale et al., 2006a). The
proposed hypothesis is that antibodies produced against GABHS crossreact against self neuronal antigens through the process of
molecular mimicry, and cause OCD.
Corresponding author. Division of Pediatric Neurology, Johns Hopkins Hospital,
Rubenstein Child Health Building, Suite 2158, 200 N. Wolfe Street, Baltimore, MD 21287,
United States. Tel.: +1 410 955 7212; fax: +1 410 614 2297.
E-mail address: hsinger@jhmi.edu (H.S. Singer).
0165-5728/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jneuroim.2009.06.015

The identication of specic antibrain antibodies reactive against

cortical or basal ganglia tissue in patients with OCD would be a strong
indicator of involvement of the autoimmune system. Two reports, using
pediatric OCD as the primary inclusion factor, have identied alterations
of immunoreactivity against striatal tissues (Dale et al., 2005; Morer
et al., 2008). In one study, Western immunoblotting in a consolidated
group of 50 OCD children (some with tic, TS, and/or other psychopathologies) showed the presence of anti-basal ganglia antibodies in
42% compared with 210% in pediatric autoimmune, neurological and
streptococcal control groups (Dale et al., 2005). Reactive immunoblot
bands were signicantly more common in the OCD cohort at 40, 45, and
60 kDa (Dale et al., 2005). In a second study, immunohistochemistry
failed to differentiate cases from controls, but immunoblotting identied
two proteins at molecular weights of 55 and 86 kDa in 7 of 32 (21 with
TS) or 22% of pre-pubertal-onset OCD patients (Morer et al., 2008).
While these studies are suggestive of a potential immune process, the
variability of results and the current controversy surrounding the
specicity of autoimmune changes in individuals with TS and PANDAS
(Church et al., 2004; Dale et al., 2006a; Singer et al., 2008, 2005, 2004),
emphasize the need for additional studies.
The aim of this study was to expand the investigation of antibrain
antibody status in pediatric OCD by using four specic pediatric
populations; OCD-only, OCD associated with PANDAS (OCD+PANDAS),
OCD plus chronic tic disorder (OCD+CTD), and Controls. We hypothesized
that the simultaneous evaluation of a) antineuronal antibodies (ANAb)
and antistreptococcal (ASO) antibodies; b) the use of three distinct

C. Gause et al. / Journal of Neuroimmunology 214 (2009) 118124

119

immunoassays; c) four brain regions; and d) and several target antigens,

might provide additional information to support or refute an autoimmune
mechanism in pediatric OCD. In addition, recognizing proposed associations between temporally-related GABHS infections and autoimmunity,
antistreptolysin O (ASO) titers were compared to antibody data and
clinical symptomatology.

used to generate a standard curve. Absorbance was measured at

450 nm using a Model 680 Microplate Reader (Bio-Rad). Microplate
Manager software (Bio-Rad) normalized the data, generated a linear
regression curve based upon standard dilutions, and calculated the
serum IgG concentrations in mg/ml. All assays were performed in
triplicate.

2. Materials and methods

2.3. Serum ASO titers

2.1. Subjects

Serum antistreptolysin O (ASO) titers were determined using the

Beckman IMMAGE 800 nephelometor in the hospital laboratories at
either the University of Minnesota or Johns Hopkins Hospital. Age
stratied values of ASO in children ages 212 years have been
published (Shet and Kaplan, 2002).

In this IRB approved study, sera from 13 OCD-only (mean age 14.1
3.1 years, range: 6.418.4, 5 girls, 8 boys) and 17 OCD+CTD children
(mean age 13.8 3.9 years, range: 7.421.4, 5 girls, 12 boys) were
obtained from patients followed by a Board-Certied Child Psychiatrist
(MG) at the Johns Hopkins Hospital. Samples were stored at 80 C.
Sera from six OCD+CTD (mean age 12.4 1.84 years, range: 10.715.1, 2
girls, 4 boys) and twenty OCD+PANDAS (mean age 11.3 1.5 years,
range: 7.913.8, 8 girls, 12 boys) subjects were available from a
multicenter Tourette Syndrome Study Group longitudinal investigation
detailed in prior manuscripts (Kurlan et al., 2008; Singer et al., 2008,
2005). These latter single-point-in-time serum samples had never been
previously evaluated. All samples were stored at 80 C and were
thawed at most only a single time prior to use in this study. All disease
subjects met their respective clinical diagnostic criteria: for OCD, DSMIV criteria (American Psychiatric Association, 2000) plus a Children's
YaleBrown Obsessive Compulsive Scale (CYBOCS) score of greater
than 16 (Scahill et al., 1997); for PANDAS, clinical criteria as established
by Swedo et al. (1998) and for chronic tic disorders (Tourette syndrome,
Chronic motor tic disorder, Chronic vocal tic disorder), clinical criteria
established by the Tourette Syndrome Classication Group (The
Tourette Syndrome Classication Study Group, 1993). Age of onset
for obsessive compulsive behaviors was as follows: OCD-only, 6.3
2.3 years; OCD+PANDAS, 6.8 2.3 years; OCD+CTD, 6.3 2.2 years.
Control subjects included 29 children (mean age 12.4 2.4 years, range:
6.218.2,17 girls,12 boys) with no lifetime personal history for either the
patient or any rst-degree relative of a tic disorder, PANDAS, TS, OCD, or
attention decit hyperactivity disorder (ADHD). No racial or ethnic
groups were excluded from this study. Laboratory personnel were
unaware of the subject grouping and all analyses were completed before
diagnostic codes were made available. Subject data is presented in
Table 1.
2.2. Serum IgG concentration measurement
Sandwich ELISA methods were used to measure serum IgG concentrations in all 85 serum samples following a previously described
methodology (Raux et al., 1999). Human IgG (Sigma-Aldrich, St. Louis,
Missouri) in serum at eight dilutions, between 3.9 and 500 ng/ml, was

Table 1
Demographic data.

Age (mean SD)

Gender
Est. age of OCD onset
CYBOCS (mean SD)
YGTSS motor
YGTSS vocal
YGTSS total tic
YGTSS total
ASO titerb
a

OCD-only
(n = 13)

OCD+PANDAS
(n = 20)

OCD+CTD
(n = 23)

Control
(n = 29)

14.1 3.1
M: 8, F: 5
6.3 2.3
26.7 7.5
00
0.8 2.8a
0.8 2.8a
0.8 2.8a
128 103

11.3 1.5
M: 12, F: 8
6.8 2.3
20.0 4.5
12.2 6.2
8.3 5.5
20.4 10.0
39.7 22.9
132 115

13.4 3.5
M: 16, F: 7
6.3 2.2
24.7 6.3
11.7 7.3
7.8 4.8
19.5 9.6
38.4 19.5
213 204

12.4 2.4
M: 12, F: 17

139 150c

One subject with a questionable history of a transient vocal tic. Clinical evaluations
were performed at the time of blood draw.
b
See Fig. 1 for box-and-whisker plot of ASO titers.
c
ASO titers were performed on 16 control samples due to limited serum availability.

2.4. Tissue preparation

Human tissue was obtained from the NICHD Brain and Tissue Bank
for Developmental Disorders at the University of Maryland, Baltimore,
Maryland. Fresh, non-frozen, human caudate, cingulate gyrus,
orbitofrontal (OFC, BA11), and dorsolateral prefrontal cortex (DLPFC,
BA9/10) were collected at autopsy from a 55-year-old Caucasian male
who died of arteriosclerotic cardiovascular disease and had no
evidence of neurological dysfunction. The postmortem interval was
approximately 8 h. A supernatant fraction for each region was
obtained by homogenizing the tissue in 0.9% normal saline (2.5 g of
tissue/10 ml of saline) containing protease inhibitors (1 g/ml of
aprotinin, 10 g/ml of leupeptin, 10 g/ml of pepstatin, and 1 mM
phenylmethylsulfonyl uoride) in a Teon-glass homogenizer on ice.
Homogenized tissue was centrifuged for 30 min at 12,000 g. The
supernatant fraction was then collected and aliquots were stored at
80 C. Protein concentrations were measured by the bicinchoninic
acid (BCA) method (Pierce, Rockford, IL).
2.5. Antibody assays
ANAb reactivity was assayed by use of three separate, distinct
methodologies; (1) immunohistochemistry, with localization of antibody binding to neuronal cells in frozen tissue sections; (2) ELISA,
measuring the amount of total antibody binding to brain homogenate;
and (3) Western immunoblotting, which identies antibodies against
homogenized and denatured proteins of a specic molecular weight,
using brain regions associated with obsessive compulsive symptoms
(orbitalfrontal cortex (Brodmann's area 11), caudate, cingulate gyrus,
and dorsolateral prefrontal cortex (Brodmann's area 9/10)), (Graybiel
and Rauch, 2000; Saxena and Rauch, 2000; Stein et al., 2000) as well
as the putative antigenic proteins pyruvate kinase M1, -enolase, enolase, and aldolase C, proposed to be associated with TS and
PANDAS (Church et al., 2004; Dale et al., 2006a).
2.5.1. Immunohistochemistry
Indirect immunouorescence was performed using a double stain
methodology with IgG and MAP2, the latter in order to localize IgG
binding to neurons. Tissue sections (20 m) were obtained from a
snap-frozen block of postmortem human striatum from a 48 year-old
female who died in an accident without head trauma. Slides were
xed in 100% methanol, washed in distilled water, and then in PBS
with agitation. Each slide was then overlaid with a blocking solution
(3% normal goat serum, 0.4% Triton X-100 in PBS) and incubated in a
dark, damp chamber at room temperature for 1 h. The slides were
again washed and then overlaid with two primary antibodies
simultaneously; 1:50 human serum and mouse anti-human microtubule associated protein 2 (MAP2) (Chemicon International, Temecula, CA) diluted 1:100 in 3% normal goat serum and 0.1% Triton X-100
in PBS. Negative controls to determine background uorescence were
run using PBS in place of serum. Slides were incubated with primary

120

C. Gause et al. / Journal of Neuroimmunology 214 (2009) 118124

antibodies for 48 h at 4 C in a dark, damp chamber. After washing,

two secondary antibodies were added; goat anti-human IgG/FITC
(Invitrogen, Eugene, OR) and Alexa Fluor 555 goat anti-mouse SFX
(Invitrogen, Eugene, OR) diluted 1:40 with 3% normal goat serum and
0.1% Triton X-100 in PBS, and incubated as above for 24 h. After
washing in PBS, MOWIOL mounting medium (Calbiochem, La Jolla,
California) was applied to each slide. Analysis was done using a Zeiss
LSM 510 laser scanning confocal microscope (Carl Zeiss Microimaging,
Thornwood, NY).
All sections were analyzed by an expert experienced neuropathologist (CP-V) who was blinded to the diagnosis. Immunoreactivity was
observed visually and scored as present or absent based on
morphology of stained section. Positive uorescence was recorded
in only nucleated cells, and colocalization was determined if the
stained neuronal process was attached to a cell body. Dilute serum
was used to reduce the degree of non-specic binding and thus
maximize sensitivity.
2.5.2. ELISA
ELISA assays were performed on all serum samples with use of
tissue protein from caudate, cingulate gyrus, prefrontal BA11, and BA9/
10 using a methodology previously described (Singer et al., 2005).
Assays were performed in sterile 96-well microtiter ELISA plates to
which 150 l of protein (2.5 g/ml) was added. The plates were
incubated overnight at 4 C, and then rinsed using washing buffer (1
PBS containing 0.05% Triton X-100). The plates were then blocked
with 5% nonfat Carnation dry milk in PBS and maintained for 90 min at
room temperature. Plates were then rinsed in washing buffer and
exposed to serum (diluted 1:250 in PBS) for 90 min at room
temperature with agitation. Secondary antibody consisted of horseradish peroxidase conjugated sheep anti-human IgG (Amersham
Biosciences, Buckinghamshire, UK) diluted 1:3000 in PBS. The amount
of antibodyantigen interaction was detected using a tetramethylbenzidine (TMB) kit purchased from Vector Laboratories (Burlingame,
CA). Freshly prepared TMB solution (TMB, buffer and hydrogen
peroxide) was allowed to incubate on the ELISA plate for 5 min,
followed by the addition of H3PO4 to quench the reaction. Immediately after TMB exposure, plates were read at optical density 450 nm
on an automated Bio-rad (Hercules, CA) Model 680 Microplate Reader.
All samples were assayed in triplicate and corrected for IgG content.
2.5.3. Western immunoblotting
Immunoblotting using caudate, cingulate gyrus, prefrontal BA11,
and BA9/10 tissue was performed on each sample. Thirty micrograms
of brain protein per sample were denatured at 100 C for 5 min,
electrophoresed in Criterion 10% TrisHCl ready-gels (Bio-Rad,
Hercules, CA), and transferred to a 0.45-m nitrocellulose membrane
at 100 V (Bio-Rad, Hercules, CA) for 80 min. The nitrocellulose was
incubated overnight at 4 C in blocking solution containing 5%
Carnation nonfat milk dissolved in TBS-T (Tris-buffered saline
containing 0.1% Tween 20). The nitrocellulose was washed three
times in TBS-T for 10 min per wash and then exposed to serum
(diluted 1:500) for 90 min at room temperature with agitation. The
nitrocellulose was washed and then exposed to secondary antibody;
horseradish peroxidase-conjugated sheep anti-human IgG (Amersham Biosciences, Buckinghamshire, UK) diluted 1:3000, and incubated for 60 min at room temperature with agitation. After washing
three times with TBS-T for 10 min per wash, the membranes were
developed with Amersham ECL reagents according to the vendor
protocol. The blots were then exposed to Denville Blue Bio Films
(Denville Scientic, Metuchen, NJ) for 60 s. Molecular weights were
estimated for each band on the basis of the distance migrated for eight
known molecular weight standards using Precision Plus Protein Dual
Color Standards (Bio-Rad, Hercules, CA). Values are presented as
2 kDa.

For immunoblotting using specic epitopes, in the place of brain

protein, 1 g of pyruvate kinase M1 (US Biological, Swampscott,
Massachusetts), -enolase (US Biological, Swampscott, Massachusetts), -enolase (Polysciences, Inc., Warrington, Pennsylvania), and
aldolase C (USB Corporation, Cleveland, Ohio) were electrophoresed
along with a constant volume of actin (US Biological), transferred to
nitrocellulose, blocked overnight, and then exposed to serum (diluted
1:500) for 90 min at room temperature, with agitation. After washing
with TBS-T, the membranes were developed with Amersham ECL
reagents as described above, stripped and re-assayed for actin content.
Western blot analyses of brain tissue homogenate and specic
antigens provided information regarding the total number of peaks
and peak height of each band. Digital image analysis and evaluation of
Western blots were performed using Quantity One software (Bio-rad,
Hercules, CA), which created densitometric data of the blots showing
the gray-intensity values (8-bit gray values) vs. Rf values. For all bands
on each blot, Quantity One generated a peak for each band, assigned
each peak a molecular weight, and determined the peak height (PH).
All peak height measurements were corrected for actin content to
avoid potential methodological discrepancies in the loading of brain
protein onto each individual gel, and for serum IgG, in order to
equalize the results based upon IgG content.
2.6. Statistical analyses
Statistical analyses were conducted in SPSS 16.0 software (SPSS
Inc., Chicago, IL). ASO titers, expressed as both Todd units and log
transformation values, were compared using one-way analysis of
variance (ANOVA) for continuous distributions and KruskalWallis,
for non-parametric conrmation if distributions were skewed. To
evaluate whether antibrain antibody proles differed among each
OCD group (OCD-only, OCD+PANDAS, and OCD+CTD) and Controls, a
four group one-way ANOVA was used for continuous data and Pearson
chi-squares, or Fisher's exact when appropriate, was used to test
categorical data frequency distributions. Western blot analyses of
brain tissue homogenate and specic antigens provided information
regarding the total number of peaks and peak height of each band.
Peak height measurements were corrected for actin content and
serum IgG concentration prior to statistical analysis using one-way
ANOVA. If the ANOVA was signicant at p b 0.01 [0.05 divided by
number of brain regions (4)], a Tukey's HSD post-hoc test, which
corrects for multiple comparisons in each ANOVA, was used to
determine specic differences between OCD groups and controls.
Specic ELISA O.D. values were calculated as total reading minus
tissue blank and analyzed using one-way ANOVA. Immunohistochemistry results were scored in a binary fashion as either present or absent
and analyzed by Pearson chi-square or Fisher's Exact test. The
frequency data, in addition to the number of individuals with a
Western blot band of a specic molecular weight, were also analyzed
using a Pearson chi-square test or Fisher's Exact test. Finally,
antineuronal antibody data and individual ASO titers were compared
to CYBOCS scores using two-tailed pairwise Pearson correlations.
3. Results
3.1. ASO
ASO values were compared using several statistical methodologies.
Fig. 1 shows a box-and-whisker plot of ASO titers among OCD groups and
controls. Top and bottom of box represent 75th and 25th percentiles,
respectively, horizontal line within each box shows the median value, and
whisker bars provide maximum and minimum data points (excluding
outliers, shown as open circles). When we analyzed continuous data using
a one-way ANOVA based on mean values, there was no signicance; OCDonly, (meanSD) n=13, 128103; OCD+PANDAS, n=20, 132115;
OCD+CTD, n=23, 213204; Controls, n=16, 139150; p=0.182;

C. Gause et al. / Journal of Neuroimmunology 214 (2009) 118124

121

Fig. 2. Composite confocal image (100) of immunohistochemical IgG staining, MAP2

staining, and colocalization.
Fig. 1. Box-and-whisker plot of ASO titers among OCD groups and controls. Top and
bottom of box represent 75th and 25th percentiles, respectively, horizontal line within
each box shows the median value, and whisker bars provide maximum and minimum
data points (excluding outliers, shown as open circles).

caudate, cingulate gyrus, and dorsolateral prefrontal cortex are

presented in Table 2. One-way ANOVA analysis indicated no differences between the groups.

df =3; F=1.668. Further, using a non-parametric method, the Kruskal

Wallis test for equality of populations showed no signicance (p=0.931,
df =3, chi-squared=0.445). The same result was obtained with a oneway ANOVA (p = 0.190; df = 3; F = 1.63) and KruskalWallis test
(p =0.233, df =3, chi-squared=4.273) for equality of ASO expressed as
log-values.

3.5. Western immunoblotting using brain homogenate

3.5.1. Bands
Western immunoblotting using postmortem tissues from caudate,
cingulate gyrus, dorsolateral prefrontal and orbitofrontal cortex
showed numerous bands against antigens in each brain region.
Pearson chi-square analysis showed that the frequency distribution
was different from expected (corrected for multiple comparisons,
p b 0.01) between groups with bands at 27 kDa (caudate, cingulate
gyrus, and DLPFC), 36 kDa (caudate), and 100 kDa (orbitofrontal
cortex and caudate), but no differences specically at 40, 45, 55, 60 and
86 kDa (Table 3). Post-hoc Fisher's Exact test analyses of the bands at
27, 36, and 100 kDa, using a multiple-comparison corrected statistical
value of p b 0.01, showed that in all instances the OCD+PANDAS group
had signicantly more bands than Controls (Table 3).

3.2. Serum IgG concentration

Measurements of serum IgG concentrations were similar among the
four groups: OCD-only, 11.7 3.6 mg/ml; OCD+PANDAS, 9.6 3.5;
OCD+CTD, 11.3 2.4, and controls, 9.9 5.2. Between group differences were not statistically signicant (ANOVA; p = 0.305; df = 3;
F = 1.2).
3.3. Immunohistochemistry

3.5.2. Peak heights

One-way ANOVA identied a difference in peak height among
groups, where n's were required to be greater than three, only at
67 kDa (OFC, caudate, cingulate gyrus, DLPFC); OFC: p b 0.001, df = 3,
F = 14.1, caudate: p b 0.001, df = 3, F = 16.4, cingulate gyrus: p b 0.001,
df = 3, F = 13.1, DLPFC: p b 0.001, df = 3, F = 13.1. Tukey post-hoc
analyses showed the OCD+PANDAS group to be signicantly different
from all other groups (p b 0.025) in all brain regions.

The number of individuals with positive IgG immunohistochemical

staining is shown in Table 2. An omnibus Pearson chi-square analysis
identied no difference in comparisons of IgG staining among groups
(p = 0.449). Comparisons of all pathological subjects (n = 69) to
controls (n = 23), were not signicant. Colocalization of IgG staining
to neurons, by use of MAP2, showed no signicant differences among
the 4 groups (p = 0.162). Fig. 2 is a composite image of immunohistochemical binding from an individual in the OCD+PANDAS cohort.

3.6. Western immunoblotting using specic antigens

3.4. ELISA
The number of subjects with reactive bands against pyruvate
kinase M1, non-neuronal enolase (-enolase), neuron-specic enolase (-enolase), and aldolase C, and their corrected peak heights, are

Mean optical density readings for serum antibodies directed

against tissue supernatant preparations from orbitofrontal cortex,
Table 2
Comparisons of immunohistochemistry and ELISA studies.
Immunohistochemistry
OCD-only
OCD+PANDAS
OCD+CTD
Control

ELISA (OD SD)

Positive IF

Colocalize (MAP2)

OFC

Caudate

Cingulate gyrus

DLPFC

3/13 (23%)
4/20 (20%)
6/23 (26%)
3/29 (10%)

3/3
3/4
5/6
1/3

0.54 0.24
0.70 0.36
0.68 0.32
0.69 0.26

0.54 0.23
0.61 0.28
0.54 0.24
0.68 0.28

0.48 0.27
0.61 0.31
0.53 0.28
0.61 0.29

0.53 0.24
0.64 0.30
0.56 0.27
0.67 0.35

(100%)
(75%)
(83%)
(33%)

Immunohistochemical studies show the number of subjects with positive IgG binding to neuronal cells, and colocalization with MAP2. Pearson Chi-square analysis identied no
signicant differences among groups. ELISA studies in different brain regions show no signicant differences between groups.
Abbreviations: IF, immunouorescence; OFC, orbitofrontal cortex; DLPFC, dorsolateral prefrontal cortex.

122

C. Gause et al. / Journal of Neuroimmunology 214 (2009) 118124

Table 3
Percent of individuals with bands at specic molecular weights in several brain regions.

OCD-only (n = 13)
OCD+PANDAS (n = 20)
OCD+CTD (n = 23)
Control (n = 29)
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value
OCD-only
OCD+PANDAS
OCD+CTD
Control
p-value

27 kDa

36 kDa

40 kDa

45 kDa

55 kDa

60 kDa

86 kDa

100 kDa

OFC

Caudate

CG

DLPFC

46%
90%
52%
66%
0.027
0%
30%
13%
14%
0.120
15%
45%
30%
48%
0.164
38%
20%
43%
34%
0.427
15%
15%
17%
17%
0.995
62%
35%
48%
52%
0.481
38%
20%
35%
34%
0.628
8%
50%**
17%
14%
0.007

46%
95%**
52%
62%
0.009
8%
50%**
9%
10%
0.001
23%
65%
48%
52%
0.131
38%
15%
43%
45%
0.147
15%
10%
22%
14%
0.748
46%
40%
57%
59%
0.567
38%
25%
39%
38%
0.748
15%
60%**
22%
17%
0.004

31%
95%**
57%
59%
0.002
8%
30%
9%
7%
0.079
23%
45%
35%
48%
0.416
31%
25%
39%
34%
0.794
15%
15%
22%
17%
0.938
38%
25%
35%
52%
0.284
38%
35%
30%
38%
0.944
0%
35%
9%
14%
0.027

31%
95%**
57%
62%
0.002
8%
40%
13%
10%
0.027
23%
50%
35%
38%
0.465
31%
25%
43%
38%
0.613
15%
15%
22%
14%
0.883
38%
30%
30%
48%
0.492
38%
40%
26%
34%
0.966
8%
30%
9%
14%
0.192

Bold p-values highlight chi-square test corrected for multiple comparisons exceeding
signicance threshold of p b 0.01.
**Indicate post-hoc comparisons test corrected for multiple comparisons that showed a
signicant difference from controls (p b 0.01). Abbreviations: OFC, orbitofrontal cortex;
CG, cingulate gyrus; DLPFC, dorsolateral prefrontal cortex.

shown in Table 4. ANOVA analysis showed no difference in either the

number of individuals with, or the mean peak height of each reactive
band.
3.7. Clinical correlations
No correlations were found between ASO titers and ELISA values
(all regions), Western blot band presence or peak height at 27, 36, 40,
45, 60, 67, and 100 kDa (all regions), putative specic antigens, and
CYBOCS scores (p N 0.05 in all comparisons).
4. Discussion
The prior identication of antibrain antibodies in children with
OCD has provided circumstantial evidence supporting the concept
that this disorder may be mediated by autoimmune mechanisms.
Results from the current study, utilizing three different immunoassays
and several different brain region-based antigenic epitopes, differ
signicantly from prior reports. ELISA, which provides an objective
semi-quantitative measurement of total antigenantibody reactivity,
using homogenized tissue, showed no difference between the three
pediatric OCD groups and Controls. Other investigators, however,

evaluating a cohort of 50 OCD-plus children (all having additional

diagnoses e.g., TS, tics, major depression, separation anxiety, phobia,
ADHD, ODD) showed elevated ELISA optical densities against striatum
as compared to control groups (Dale et al., 2005). Whether
methodological differences explain the variability of results is
unknown. ELISA with serum has shown consistently higher optical
densities when using fresh brain as compared to frozen brain
homogenates (Rippel et al., 2005). Additionally, in this study ELISA
results were corrected for IgG content.
Immunohistochemistry has a long history of antineuronal antibody detection and for some investigators represents the methodology of choice for identication, i.e., epitopes have not been denatured
by homogenization or boiling nor potentially lost in centrifugation. In
this study, no group differences in reactivity were identied in IgG
binding to human striatum, similar to the report by Morer et al.
(2008). Further, in individuals with positive reactivity, colocalization
of neuronal staining using MAP2 failed to identify a difference in
frequency distribution among groups. In contrast, other studies have
shown positive staining to basal ganglia in two-thirds of OCD
+PANDAS subjects (Pavone et al., 2004) and reactivity was associated
with OC symptoms and possibly tics (Kiessling et al., 1993). A potential
explanation for immunohistochemical discrepancies is the use of
different microscope detection systems and dilutions of sera. We favor
the use of a highly sensitive laser scanning confocal microscope and
use of a lower dilution of serum (Morris et al., 2009).
Western blot analyses separate potential epitopes, homogenized
and boiled, based on their molecular weight, and can identify whether
a specic reactive band or its corrected peak height occurs more
frequently in a specic OCD population. No differences in either the
number of bands or peak heights were identied against previously
proposed epitopes at 40, 45, 55, 60, or 86 kDa (Dale et al., 2005; Morer
et al., 2008) or when using the putative specic antigens and enolase, aldolase C, and pyruvate kinase M1 as the epitope (Dale et al.,
2005). Analyses did, however, show differences for the number of
bands at 27 kDa (caudate, cingulate gyrus, and DLPFC), 36 kDa
(caudate), and 100 kDa (caudate and OFC), with post hoc evaluation
identifying that OCD+PANDAS had more bands than controls.
Whether these changes have relevance to the presence of OCD is
unknown. None of the bands reported in this study have been
identied in prior studies of children with TS, PANDAS, or OCD. For
example, prior studies have shown reactive bands in children with:
PANDAS, at 183, 60, 45, and 40 kDa (Church et al., 2003; Singer et al.,
2004); TS, at 120, 83, 67, 60, 45, and 40 kDa (Church et al., 2003;
Singer et al., 1998; Wendlandt et al., 2001; Yeh et al., 2006); and TS+
OCD, at 83 kDa (Pavone et al., 2004). Differences in peak height were
identied among groups only at 67 kDa, with OCD+PANDAS being
greater than all others at OFC, caudate, cingulate gyrus, and DLPFC. In a
prior study, children with TS had a greater number of bands at 67 kDa
than controls (peak heights not assessed), but their presence was
not associated with the diagnosis of OCD (Singer et al., 1998). It is

Table 4
Antibodies against putative specic antigens.

OCD-only (n = 13)
Corrected peak height
OCD+PANDAS (n = 20)
Corrected peak height
OCD+CTD (n = 23)
Corrected peak height
Control (n = 29)
Corrected peak height

Aldolase C

-enolase

-enolase

PK M1

77%
0.048 0.023
75%
0.051 0.033
74%
0.038 0.026
62%
0.059 0.037

77%
0.058 0.016
95%
0.082 0.052
70%
0.051 0.031
90%
0.102 0.091

31%
0.050 0.005
15%
0.062 0.013
35%
0.034 0.015
7%
0.065 0.006

31%
0.058 0.021
40%
0.053 0.011
35%
0.045 0.017
45%
0.094 0.086

The number of subjects with; and peak height (mean standard deviation) of Western
blot bands. Each band represents an antibodyantigen interaction between serum
autoantibodies and a putative antigenic epitope. No signicant differences were
observed. Abbreviations: PK M1, pyruvate kinase M1.

C. Gause et al. / Journal of Neuroimmunology 214 (2009) 118124

important to consider that the mere presence of autoantibodies, or

their peak height, does not convey a causal association. Antibrain
antibodies have been identied in normal individuals (Archelos and
Hartung, 2000; Lacroix-Desmazes et al., 1998; Singer et al., 2005), may
be the result of secondary damage or a non-specic epiphenomenon
(Bornstein et al., 2001), do not correlate with clinical exacerbations in
subjects with PANDAS (Singer et al., 2008), and differences may reect
methodological variations between laboratories (Rippel et al., 2005).
In formal comparisons between ECL and colorimetric detection
methods, results have suggested that ECL may allow for more sensitive
quantication (Rippel et al., 2005). Differences in reactivity to
pyruvate kinase M1 and aldolase C could be related to our use of a
commercial preparation derived from rabbit muscle having only 93
and 79% homology, respectively, with human brain isoforms (Dale et
al., 2006b).
Support for the association of GABHS with OCD is based on reports
of the exacerbation of symptoms with streptococcal infections
(Giulino et al., 2002; Mell et al., 2005; Murphy and Pichichero,
2002; Murphy et al., 2004; Swedo et al., 1998) and on the presence of
elevated ASO titers in subjects with OCD and tics (Cardona and Oreci,
2001; Murphy et al., 2004). Our data identied no difference in ASO
titers among children with OCD-only, OCD+PANDAS, OCD+CTD, and
Controls and no correlation between ASO titer and OC severity, as
measured by the CYBOCS.
Originating from a cross-sectional sample, the data cannot be used
to clarify conicting longitudinal reports on a possible association
between streptococcal infection and OC symptom exacerbations (Luo
et al., 2004; Mell et al., 2005; Murphy et al., 2004). In a serial
multicenter investigation following 40 subjects with PANDAS and 40
subjects with Tourette syndrome, only 5 of 64 exacerbations were
temporally associated (within 4 weeks) with a group A -hemolytic
streptococcus infection, and all occurred in PANDAS cases (Kurlan
et al., 2008). Our data also demonstrate no association between ASO
titers and reactive bands identied on Western Blot analysis, in
keeping with prior reports in OCD and TS subjects (Morer et al., 2008;
Singer et al., 1998). In contrast, in a UK study of subjects with reactive
bands at 40, 45, and 60 kDa, the data suggest that elevated binding to
anti-basal ganglia (neuronal) antibodies is more likely to occur in
individuals with the diagnosis of OCD+TS and higher ASO titers (Dale
et al., 2005). It has been noted that antistreptococcal antibody titers
are frequently misinterpreted based on the failure to recognize that
the normal levels of these antibodies are higher in school-aged
children than in adults (Gerber et al., 2009; Kaplan et al., 1998).
This study has several limitations including the inclusion of only a
moderate number of subjects in each OCD category, varying age
ranges, a lack of consideration of illness duration, and the failure to
assess other antigenic targets such as tubulin and lysoganglioside GM1
(identied in SC) (Kirvan et al., 2007, 2006) and serum cytokines,
suggested to be associated with OCD (Mittleman et al., 1997).
Importantly, it is emphasized that the establishment of an autoimmune disorder requires not only the identication of autoantibodies, but the presence of immunoglobulins at the pathological site, a
positive response to immunomodulatory therapy, the induction of
symptoms with autoantigens, and the ability to passively transfer the
disorder to animal models with the induction of behavioral symptoms
(Archelos and Hartung, 2000). Additional studies are required to
achieve this requirement and to clarify the role of autoimmunity in
OCD.

Acknowledgements
We thank the NICHD Brain and Tissue Bank for Developmental
Disorders at the University of Maryland, Baltimore, Maryland for
supplying fresh, non-frozen postmortem brain tissue and Mathew
Pollard for his assistance in the laboratory.

123

References
American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental
Disorders: DSM-IV-TR (4th ed.) (text revision). Washington, D.C.
Archelos, J.J., Hartung, H.P., 2000. Pathogenetic role of autoantibodies in neurological
diseases. Trends Neurosci. 23, 317327.
Bornstein, N.M., Aronovich, B., Korczyn, A.D., Shavit, S., Michaelson, D.M., Chapman, J.,
2001. Antibodies to brain antigens following stroke. Neurology 56, 529530.
Cardona, F., Oreci, G., 2001. Group A streptococcal infections and tic disorders in an
Italian pediatric population. J. Pediatr. 138, 7175.
Church, A.J., Cardoso, F., Dale, R.C., Lees, A.J., Thompson, E.J., Giovannoni, G., 2002. Antibasal ganglia antibodies in acute and persistent Sydenham's chorea. Neurology 59,
227231.
Church, A.J., Dale, R.C., Lees, A.J., Giovannoni, G., Robertson, M.M., 2003. Tourette's
syndrome: a cross sectional study to examine the PANDAS hypothesis. J. Neurol.
Neurosurg. Psychiatry 74, 602607.
Church, A.J., Dale, R.C., Giovannoni, G., 2004. Anti-basal ganglia antibodies: a possible
diagnostic utility in idiopathic movement disorders? Arch. Dis. Child. 89, 611614.
Dale, R.C., Heyman, I., Giovannoni, G., Church, A.W., 2005. Incidence of anti-brain
antibodies in children with obsessive-compulsive disorder. Br. J. Psychiatry 187,
314319.
Dale, R.C., Candler, P.M., Church, A.J., Wait, R., Pocock, J.M., Giovannoni, G., 2006a.
Neuronal surface glycolytic enzymes are autoantigen targets in post-streptococcal
autoimmune CNS disease. J. Neuroimmunol. 172, 187197.
Dale, R.C., Church, A.J., Candler, P.M., Chapman, M., Martino, D., Giovannoni, G., 2006b.
Serum autoantibodies do not differentiate PANDAS and Tourette syndrome from
controls. Neurology 66, 1612 author reply 1612.
de Mathis, M.A., do Rosario, M.C., Diniz, J.B., Torres, A.R., Shavitt, R.G., Ferrao, Y.A.,
Fossaluza, V., de Braganca Pereira, C.A., Miguel, E.C., 2008. Obsessivecompulsive
disorder: inuence of age at onset on comorbidity patterns. Eur. Psychiatry 23,
187194.
Gerber, M.A., Baltimore, R.S., Eaton, C.B., Gewitz, M., Rowley, A.H., Shulman, S.T., Taubert,
K.A., 2009. Prevention of rheumatic fever and diagnosis and treatment of acute
streptococcal pharyngitis. A Scientic Statement From the American Heart
Association Rheumatic Fever, Endocarditis, and Kawasaki Disease Committee of
the Council on Cardiovascular Disease in the Young, the Interdisciplinary Council on
Functional Genomics and Translational Biology, and the Interdisciplinary Council
on Quality of Care and Outcomes Research. Circulation 119, 15411551.
Giulino, L., Gammon, P., Sullivan, K., Franklin, M., Foa, E., Maid, R., March, J.S., 2002. Is
parental report of upper respiratory infection at the onset of obsessivecompulsive
disorder suggestive of pediatric autoimmune neuropsychiatric disorder associated
with streptococcal infection? J. Child Adolesc. Psychopharmacol. 12, 157164.
Graybiel, A.M., Rauch, S.L., 2000. Toward a neurobiology of obsessivecompulsive
disorder. Neuron 28, 343347.
Kaplan, E.L., Rothermel, C.D., Johnson, D.R., 1998. Antistreptolysin O and antideoxyribonuclease B titers: normal values for children ages 2 to 12 in the United
States. Pediatrics. 101, 8688.
Kiessling, L.S., Marcotte, A.C., Culpepper, L., 1993. Antineuronal antibodies in movement
disorders. Pediatrics 92, 3943.
Kirvan, C.A., Swedo, S.E., Kurahara, D., Cunningham, M.W., 2006. Streptococcal mimicry
and antibody-mediated cell signaling in the pathogenesis of Sydenham's chorea.
Autoimmunity 39, 2129.
Kirvan, C.A., Cox, C.J., Swedo, S.E., Cunningham, M.W., 2007. Tubulin is a neuronal target
of autoantibodies in Sydenham's chorea. J. Immunol. 178, 74127421.
Kurlan, R., Johnson, D., Kaplan, E.L., 2008. Streptococcal infection and exacerbations of
childhood tics and obsessivecompulsive symptoms: a prospective blinded cohort
study. Pediatrics 121, 11881197.
Lacroix-Desmazes, S., Kaveri, S.V., Mouthon, L., Ayouba, A., Malanchere, E., Coutinho, A.,
Kazatchkine, M.D., 1998. Self-reactive antibodies (natural autoantibodies) in
healthy individuals. J. Immunol. Methods 216, 117137.
Luo, F., Leckman, J.F., Katsovich, L., Findley, D., Grantz, H., Tucker, D.M., Lombroso, P.J.,
King, R.A., Bessen, D.E., 2004. Prospective longitudinal study of children with tic
disorders and/or obsessivecompulsive disorder: relationship of symptom exacerbations to newly acquired streptococcal infections. Pediatrics 113, e578585.
Mell, L.K., Davis, R.L., Owens, D., 2005. Association between streptococcal infection and
obsessivecompulsive disorder, Tourette's syndrome, and tic disorder. Pediatrics
116, 5660.
Mercadante, M.T., Busatto, G.F., Lombroso, P.J., Prado, L., Rosario-Campos, M.C., do Valle, R.,
Marques-Dias, M.J., Kiss, M.H., Leckman, J.F., Miguel, E.C., 2000. The psychiatric
symptoms of rheumatic fever. Am. J. Psychiatry 157, 20362038.
Mittleman, B.B., Castellanos, F.X., Jacobsen, L.K., Rapoport, J.L., Swedo, S.E., Shearer, G.M.,
1997. Cerebrospinal uid cytokines in pediatric neuropsychiatric disease. J. Immunol.
159, 29942999.
Morer, A., Lazaro, L., Sabater, L., Massana, J., Castro, J., Graus, F., 2008. Antineuronal
antibodies in a group of children with obsessivecompulsive disorder and Tourette
syndrome. J. Psychiatr. Res. 42, 6468.
Morris, C.M., Pardo-Villamizar, C., Gause, C.D., Singer, H.S., 2009. Serum autoantibodies
measured by immunouorescence conrm a failure to differentiate PANDAS and
Tourette syndrome from controls. J. Neurol. Sci. 276, 4548.
Murphy, M.L., Pichichero, M.E., 2002. Prospective identication and treatment of children
with pediatric autoimmune neuropsychiatric disorder associated with group A
streptococcal infection (PANDAS). Arch. Pediatr. Adolesc. Med. 156, 356361.
Murphy, T.K., Sajid, M., Soto, O., Shapira, N., Edge, P., Yang, M., Lewis, M.H., Goodman, W.K.,
2004. Detecting pediatric autoimmune neuropsychiatric disorders associated with
streptococcus in children with obsessivecompulsive disorder and tics. Biol. Psychiatry
55, 6168.

124

C. Gause et al. / Journal of Neuroimmunology 214 (2009) 118124

Murphy, T.K., Sajid, M.W., Goodman, W.K., 2006. Immunology of obsessivecompulsive

disorder. Psychiatr. Clin. North Am. 29, 445469.
Pavone, P., Bianchini, R., Parano, E., Incorpora, G., Rizzo, R., Mazzone, L., Triletti, R.R.,
2004. Anti-brain antibodies in PANDAS versus uncomplicated streptococcal
infection. Pediatr. Neurol. 30, 107110.
Raux, M., Finkielsztejn, L., Salmon-Ceron, D., Bouchez, H., Excler, J.L., Dulioust, E., Grouin,
J.M., Sicard, D., Blondeau, C., 1999. Development and standardization of methods to
evaluate the antibody response to an HIV-1 candidate vaccine in secretions and sera
of seronegative vaccine recipients. J. Immunol. Methods 222, 111124.
Rippel, C.A., Hong, J.J., Yoon, D.Y., Williams, P.N., Singer, H.S., Rippel, C.A., Hong, J.J., Yoon,
D.Y., Williams, P.N., Singer, H.S., 2005. Methodological factors affect the measurement of anti-basal ganglia antibodies. Ann. Clin. Lab. Sci. 35.
Saxena, S., Rauch, S.L., 2000. Functional neuroimaging and the neuroanatomy of
obsessivecompulsive disorder. Psychiatr. Clin. North Am. 23, 563586.
Scahill, L., Riddle, M.A., McSwiggin-Hardin, M., Ort, S.I., King, R.A., Goodman, W.K.,
Cicchetti, D., Leckman, J.F., 1997. Children's YaleBrown Obsessive Compulsive
Scale: reliability and validity. J. Am. Acad. Child Adolesc. Psychiatry 36, 844852.
Shet, A., Kaplan, E.L., 2002. Clinical use and interpretation of group A streptococcal
antibody tests: a practical approach for the pediatrician or primary care physician.
Pediatr. Infect. Dis. J. 21, 420426.
Singer, H.S., Giuliano, J.D., Hansen, B.H., Hallett, J.J., Laurino, J.P., Benson, M., Kiessling, L.S.,
1998. Antibodies against human putamen in children with Tourette syndrome.
Neurology 50, 16181624.
Singer, H.S., Loiselle, C.R., Lee, O., Minzer, K., Swedo, S., Grus, F.H., 2004. Anti-basal
ganglia antibodies in PANDAS. Mov. Disord. 19, 406415.

Singer, H.S., Hong, J.J., Yoon, D.Y., Williams, P.N., 2005. Serum autoantibodies do not
differentiate PANDAS and Tourette syndrome from controls. Neurology 65,
17011707.
Singer, H.S., Gause, C., Morris, C., Lopez, P., 2008. Serial immune markers do not
correlate with clinical exacerbations in pediatric autoimmune neuropsychiatric
disorders associated with streptococcal infections. Pediatrics 121, 11981205.
Stein, D.J., Goodman, W.K., Rauch, S.L., 2000. The cognitiveaffective neuroscience of
obsessivecompulsive disorder. Curr. Psychiatry Rep. 2, 341346.
Swedo, S.E., 1994. Sydenham's chorea. A model for childhood autoimmune neuropsychiatric disorders. Jama 272, 17881791.
Swedo, S.E., Rapoport, J.L., Cheslow, D.L., Leonard, H.L., Ayoub, E.M., Hosier, D.M., Wald,
E.R., 1989. High prevalence of obsessivecompulsive symptoms in patients with
Sydenham's chorea. Am. J. Psychiatry 146, 246249.
Swedo, S.E., Leonard, H.L., Garvey, M., Mittleman, B., Allen, A.J., Perlmutter, S., Lougee, L.,
Dow, S., Zamkoff, J., Dubbert, B.K., 1998. Pediatric autoimmune neuropsychiatric
disorders associated with streptococcal infections: clinical description of the rst
50 cases. Am. J. Psychiatry 155, 264271.
The Tourette Syndrome Classication Study Group, 1993. Denitions and classication
of tic disorders. Arch. Neurol. 50, 10131016.
Wendlandt, J.T., Grus, F.H., Hansen, B.H., Singer, H.S., 2001. Striatal antibodies in children
with Tourette's syndrome: multivariate discriminant analysis of IgG repertoires.
J. Neuroimmunol. 119, 106113.
Yeh, C.B., Wu, C.H., Tsung, H.C., Chen, C.W., Shyu, J.F., Leckman, J.F., 2006. Antineural
antibody in patients with Tourette's syndrome and their family members. J. Biomed.
Sci. 13, 101112.