An Autistic Endophenotype Results in Complex Immune

Dysfunction in Healthy Siblings of Autistic Children

Marina Saresella, Ivana Marventano, Franca Rosa Guerini, Roberta Mancuso, Lara Ceresa,
Milena Zanzottera, Beatrice Rusconi, Emanuela Maggioni, Carmine Tinelli, and Mario Clerici
Background: Endophenotypes are simple biological aspects of a disease that can be observed in unaffected relatives at a higher rate than
in the general population; an autism endophenotype justifies the observation that a mild reduction in ideational fluency and nonverbal
generativity might be observed in healthy, unaffected relatives of children with autism. Because it is becoming apparent that autism is
associated with given alleles encoding within the human leukocyte antigens region, a region of pivotal importance in immunity, we
examined whether the autism endophenotype would extend its effects on the immune system.
Methods: Multiple immune parameters were analyzed in autistic children (AC) (n 20), their siblings (HSAC) (n 15), and age- and
gender-comparable healthy control subjects (HC) (n 20) without any familiarity for autism.
Results: The immune profiles of HSAC were significantly more similar to those of their autistic siblings than to what was observed in HC.
Thus, in AC and HSAC compared with HC: 1) proinflammatory and interleukin-10 producing immune cells were augmented (p .01 in both
comparisons); 2) CD8 nave (CD45RA/CCR7) T lymphocytes were increased (p .0001 and p .001); and 3) CD8 effector memory
(CD45RA/CCR7) (p .0001 and p .03) as well as CD4 terminally differentiated (CD45RA/CCR7) (p .05 in both comparisons)
lymphocytes were diminished. Serum autoantibodies (GM1) could be detected in 10% of AC children alone.
Conclusions: Results of this pilot study indicate that a complex immune dysfunction is present both in autistic children and in their
non-autistic siblings and show the presence of an autism endophenotype that expands its effects on immunologic functions.
Key Words: Autism, autistic endophenotype, cytokines, immunology, T lymphocytes, thymus

utism is a disorder of childhood associated with behavioral impairments of communication, social interaction,
learning and thinking skills, and with repetitive and
auto-aggressive behaviors (1). The etiology of autism is still
unknown, but several elementsincluding genetic, infectious,
and immunologic factorsare suspected to play a role in this
disease (2). Systemic immunologic aberrations in autism have
been linked to autoimmunity (3). Thus, autoantibodies directed
toward central nervous system (CNS) proteins (4 8) and elevated serum titres of antibodies (9,10) were described in children
affected by autism (AC). More subtle immune impairments,
including alterations of mitogen-induced proliferation (11,12),
reduced number of total lymphocytes (13), impairments of the
CD4/CD8 T cells ratio (14 18), defective T cell activation (12,17),
and reduced natural killer (NK) cytotoxicity, have also been
observed in autistic children (19). Finally, alterations in the
Th1/Th2 cytokine balance (20) and a dysregulation in apoptosis
mechanisms (21) can also be present in autistic patients.
An autism endophenotype was recently hypothesized to
justify the observations that relatives of autistic children can
present autistic traits, including delayed verbal, cognitive, and

From the Laboratory of Molecular Medicine and Biotechnology (MS, IM, FRG,
RM, LC, MZ, MC); Department of Neuropsychiatry (EM), Don C. Gnocchi
ONLUS Foundation IRCCS-S.M. Nascente; Department of Neuropsychiatry (BR), Sacra Famiglia Institute; Chair of Immunology (MC), Department
of Biomedical Sciences and Technologies, University of Milano, Milano;
and the Clinical Epidemiology and Biometric Unit (CT), IRCCS Policlinico
S. Matteo Foundation, Pavia, Italy.
Address correspondence to Marina Saresella, Ph.D., Laboratory of Molecular
Medicine and Biotechnology, Don C. Gnocchi ONLUS Foundation IRCCS,
Via Capecelatro 66, Milan 20148, Italy; E-mail: msaresella@dongnocchi.it.
Received Jan 9, 2009; revised Jun 25, 2009; accepted Jun 26, 2009.

0006-3223/09/$36.00
doi:10.1016/j.biopsych.2009.06.020

motor development (2225). The association between particular

human leukocyte antigen (HLA) alleles and autism has been
repeatedly shown (26 28). The HLA genes are of pivotal importance in immunity, and immune cells and immune proteins can
modulate brain functions, affecting cognitive and emotional
processing. The autism endophenotype could therefore also
include immune alterations that might be associated with a
modification of cognitive and emotional processes. To examine
whether immune impairments are present in healthy, nonautistic siblings of autistic children, we performed an in-depth
analysis of a number of immune parameters. Results indicate that
autism is indeed associated with a complex pattern of immune
dysfunction that affects multiple immune cell subsets; notably,
this immune impairment is also clearly observed in non-autistic
siblings of autistic children.

Methods and Materials

Patients and Control Subjects
Fifty-five children were enrolled in the study. Twenty children
(median age 13 years; range 517 years; 6 female and 14
male subjects) had a diagnosis of autism (AC) according to
DSM-IV criteria. These patients were characterized by normal
growth parameters, blood pressure, liver function, and ECG.
None of the patients had undergone any therapy, including
psychotropic drugs, for at least 1 month before the study. Fifteen
healthy, unaffected siblings of the AC patients were also enrolled
in the study (HSAC) (median age 15 years; range 316 years;
9 female and 6 male subjects). Sub-threshold levels of autistic
social impairment as evaluated by the Social Responsiveness
Scale score were present in the HSAC. Anamnestic data were
analyzed for all children; both groups were immunized according to the vaccination schedules of the Italian Ministry of Health;
no major infectious or inflammatory event was recorded in the
clinical history of any AC or HSAC children. A third group of
individuals included 20 healthy control subjects (HC) (median
age 11 years; range 315 years; 9 female and 11 male
BIOL PSYCHIATRY 2009;66:978 984
2009 Society of Biological Psychiatry

M. Saresella et al.

BIOL PSYCHIATRY 2009;66:978 984 979

Figure 1. Cytokine producing CD4 T lymphocytes in

autistic patients (AC), their healthy siblings (HSAC),
and healthy children (HC) without familiarity for autism. Mitogen-stimulated results (background subtracted) are presented. The boxes stretch from the
25th to the 75th percentile; the lines across the boxes
indicate the median values; the lines stretching from
the boxes indicate extreme values. Statistical significance is shown.

subjects) without familiarity for autism. To avoid the influence of

pollens and minimize that of viral and bacterial infection, all the
samples were collected in summer 2008; additionally, all samples
were collected at the same time of the day (early morning).
This study received approval from the Ethics Committee of
the Don Gnocchi Foundation; written informed consent was

obtained by the individuals enrolled in the study or by their legal

guardian.
Blood Sample Collection
Whole blood was collected in vacutainer tubes containing
ethylenediamine tetra-acetic acid (EDTA) (Becton Dickinson,

Figure 2. Cytokine producing CD8 T lymphocytes in AC, HSAC, and HC without familiarity for autism. Mitogen-stimulated results (background subtracted)
are presented. The boxes stretch from the 25th to the 75th percentile; the lines across the boxes indicate the median values; the lines stretching from the boxes
indicate extreme values. Statistical significance is shown. IL, interleukin; IFN, interferon; TNF, tumor necrosis factor; other abbreviations as in Figure 1.

www.sobp.org/journal

980 BIOL PSYCHIATRY 2009;66:978 984

Rutherford, New Jersey). Peripheral blood mononuclear cells
(PBMC) were separated on lymphocyte separation medium
(Organon Teknika, Durham, North Carolina) and washed twice
in phosphate-buffered saline (PBS); viable leukocytes were
determined by trypan blue exclusion. Serum samples were
stored at 20C before use.
Intracellular Cytokines and Perforin and Granzyme
Production
Freshly isolated PBMC (1 106) were resuspended in medium alone or stimulated with SEB (5 g/mL) (Sigma, St. Louis,
Missouri) anti CD28 (2 g/mL) (Sigma) or, to measure perforin
and granzymes, interleukin (IL)-2 (5000 U/mL) (Sigma). Culture
tubes were incubated at 37C in a humidified 5% carbon dioxide
atmosphere for 18 hours; Brefeldin A (Sigma) (10 mg/mL) was
added to the cultures in the final 5 hours.
Immunophenotypic Analysis
Lymphocyte subsets were evaluated with a Cytomics FC-500
flow cytometer (Beckman Coulter, Fullerton, California) on 100
L of EDTA peripheral blood incubated for 30 min at 4C with
fluorochrome-labeled monoclonal antibodies. Erythrocyte lysis
was obtained with the Immuno-Prep Epics Kit and Q-Prep Work
Station (Beckman Coulter). For each analysis, 20,000 events were
acquired and gated on CD4, CD8, or CD14 expression and side
scatter properties. Samples were first run with single fluorochrome-stained preparation for color compensation.
Monoclonal Antibodies
The following monoclonal antibodies (mAbs) were used:
anti-human CD28 (CD28.2), anti-Phycoerythrin-Cyanin-7 (PC7)labeled anti-CD4 (clone SFCI12T4D11), Phycoerythrin-Cyanin-5
(PC5)-labeled anti-CD8 (clone B9.11), Fluorescein isothiocyanate
(FITC)-labeled anti-CD45RA (clone 2H4LDH11LDB9), Phycoerythrin (PE)-labeled anti-CD3 (UCHT1), FITC-labeled anti-CD16
(3G8), PE-labeled anti-CD38 (T16) (all Mabs isotype m immunoglobulin G [IgG]1), FITC-labeled anti-CD45RO (UCHL1) (Beckman Coulter), and PE-labeled anti-CCR7 (clone 150,503) (both
m IgG2a) (R&D System, Minneapolis, Minnesota). Anti-IL2-FITC
(MQ1-17H12) (rat-IgG2a), anti-tumor necrosis factor (TNF)FITC(MP9-20A4), anti-IL10-PE (JES3-9D7) (rat IgG1), anti-interferon (IFN)-PE(B27), anti-IL12-PE (C8.6), anti-IL1-FITC(ILB1),
anti-IL6-PE (B-E8) (m IgG1) (Invitrogen-Caltag Lab, Carlsbad,
California), FITC-conjugated anti-perforin (. G9) (m IgG2b), and
anti-granzyme-PE (GB11) (m IgG1) (Becton Dickinson Biosciences, San Jose, California) as well as isotype-matched mouse
mAbs were also used.
Immunofluorescent Staining
Stimulated PBMC were washed in PBS, split in tubes, and
stained with CD4, CD8, and CD14 mAb for 30 min at 4C in the
dark. The PBMC were then washed, and the intracellular staining
of cytokine, perforin, and granzymes was conducted with the FIX
and PERM kit (Caltag-Invitrogen).
Autoantibodies
Sera were assayed by enzyme-linked immunosorbent assay
(ELISA) for antibodies to sulfatide, monoganglioside GM1, asialo
monoganglioside GM1 (Sigma Aldrich, Milano, Italy) and tetrasialoganglioside GQ1b (Alexi Biochemicals, Lausen, Switzerland) as
previously described (29). Antimyelin-associated glycoprotein
(anti-MAG) antibodies were evaluated with a commercial kit
(Buhlmann Laboratories, Schonenbuch, Switzerland), whereas
antibodies against HuD, Ri, Yo, and amphiphysin were assayed
www.sobp.org/journal

M. Saresella et al.
with an immunoassay kit in a dot-blot format (Milenia Biotec,
Bad Nauheim, Germany).
Virological Analysis
Serologic analyses were performed with ELISA kits for Measles virus, herpes simplex virus (HSV) 1/2, varicella zoster virus
(VZV), Parvovirus-B19, (Genzyme Virotech, Cambridge, Massachusetts), Toxoplasma (AxSYM, Abbott Park, Illinois), and by
immunofluorescent antibody (IFA) test for Epstein Barr Virus
Table 1. Cytokines Producing CD4 and CD8 T Lymphocytes and
CD14 Monocytes and Perforin or Granzyme Producing CD8 T
Lymphocytes in AC, HSAC, and HC
Cell Populations

AC
(n 20)

HSAC
(n 15)

HC
(n 20)

CD4 IL2

.3
.15.76
.6
.3.71
.8b
.51.4
3.4c
2.84.6
1.3c
.72.8
.3
.13.9
0
0.05
.2c
.2.3
.9b
.41.3
.4c
.3.7
3.9c
2.15.5
.3c
.21.1
.57
.35.66
.02
0.07
11.8c
6.514.2
76.5
7378.4
5.8c
2.76.8
4.1
2.55.7
40.7c
20.575.7
13.9
12.722.3
.4
.24.91

.4
.2.5
.3a
.3.5
1.5
.91.9
.1a,c
.1.2
0a,c
0.1
.2
0.4
0
00
0a,c
00
1
.62.1
1c
.61.3
0a,c
.1.1
0a,c
0.1
.5
.16.8
0
0.07
21.8c
1538.8
78.7a
74.685.3
2.1a,c
1.32.85
2.2
1.53.6
77.5a,c
67.584
15
12.3821.68
.365
.13.98

CD4 IFN
CD4 TNF
CD4 IL6
CD4 IL10
CD4 IL12
CD4 IL1
CD8 IL2
CD8 IFN-
CD8 TNF
CD8 IL6
CD8 IL10
CD8 IL12
CD8 IL1
CD14 TNF
CD14 IL6
CD14 IL10
CD14 IL12
CD14 IL1
CD8 PERFORIN
CD8 GRANZYME

.51.45
.7a
.531.35
1.5b
.92.5
2.3a
.63.2
1.4a
.51.9
.5
.291.45
0
0.01
.1a
.1.3
1.6b
1.22.1
.8
.31.3
1.7a
.83.1
.3a
.2.7
.7
.32.15
0
00
16.9
4.934.1
67a
47.976.3
4.4a
1.77.8
3.9
1.37.3
36.5a
1157
13.1
7.8718
.37
.051.4

p .0021a
p .04b
p .0001a,c
p .0001a,c

p .015a
p .0001c
p .01b
p .01c
p .0001a,c
p .0001a,c

p .008c
p .005a
p .006a
p .01c
p .0001a
p .007c

Median and interquartile ranges (% of cells) are shown. All p values are
Bonferroni-corrected.
IL, interleukin; IFN, interferon; TNF, tumor necrosis factor.
a
Children with autism (AC) versus healthy control subjects (HC).
b
AC versus siblings without autism (HSAC).
c
HSAC versus HC.

M. Saresella et al.
Viral Capsid Antigen (EBV-VCA) (Gull Laboratories, Salt Lake
City, Utah).
Statistical Analysis
Quantitative data were not normally distributed (Shapiro
Wilk test) and are thus summarized as median and Interquartile
Range (IQR) (25 and 75 percentile). Comparisons between
groups were analyzed to evaluate immunologic differences.
KruskalWallis analysis of variance was performed for each
variable; Bonferroni correction was applied to the results. The
p values and all tests are two-sided. Data analysis was performed
with the Stata statistical package (version 9; Stata Corporation,
College Station, Texas).

Results
Cytokine-Producing CD4 and CD8 T Lymphocytes
The IL-1-, IL-2-, IL-6-, IL-10-, IL-12-, TNF-, and IFN-producing T lymphocytes were measured both in unstimulated
cultures and upon stimulation with SEB antiCD28. No significant differences were detected in the parameters analyzed in
unstimulated conditions. In contrast with these results, IL-6 and
IL-10-producing SEB antiCD28-stimulated CD4 and CD8 T
lymphocytes were greatly increased both in AC and HSAC
compared with HC (p .0001 for all comparisons), and IL-2producing CD8 T lymphocytes were significantly augmented in
AC and HSAC compared with HC (AC vs. HC p .0015; HSAC vs.
HC p .0001).
Additional results showed that TNF- and IFN--producing
CD4 T lymphocytes were significantly augmented in AC compared with HSAC and with HC (p .05 in all cases), and
IFN--producing CD8 T cells were increased in AC (p .01)
compared with HSAC.
The highest percentage of TNF-producing CD8 T cells was
seen, somewhat unexpectedly, in HC (vs. HSAC p .013).
Finally, no differences were detected in either CD4 or CD8

BIOL PSYCHIATRY 2009;66:978 984 981

IL-12- and IL-1 producing T cells when AC, HSAC, and HC were
compared. These results are shown in Figures 1 and 2 and in
Table 1.
Cytokine-Producing CD14 Cells
The SEB antiCD28-stimulated IL-10-producing CD14
cells were augmented in AC (p .006) compared with HC;
IL-1- and IL-6-producing CD14 cells were reduced in AC
compared with HC (p .0001 and p .005, respectively). No
clear patterns were seen in HSAC, in whom, nevertheless,
IL-10-producing cells were increased (p .012) and IL-1and TNF-producing cells were diminished (p .007 and p
.008, respectively) compared with HC (Figure 3, Table 1).
Immunophenotipical Analysis
The CD4 T cells were augmented, and B lymphocytes were
diminished, albeit not significantly, in AC and HSAC compared
with HC (Table 2). A more complex pattern of alterations was
detected when memory and nave T lymphocyte subsets were
investigated. Thus, in AC and HSAC compared with HC: 1) nave
(CCR7/CD45RA) CD8 T lymphocytes were increased (p
.0001 and p .001, respectively); 2) effector memory (EM)
(CCR7/CD45RA) CD8 T cells were reduced (p .0001 and
p .03, respectively); and 3) terminally differentiated (TD)
(CCR7/CD45RA) CD4 T lymphocytes were reduced (p
.013 and p .002, respectively). These results, summarized in
Figure 4 and Table 2, indicate that post-thymic differentiation
pathways are profoundly altered in autistic children and in their
unaffected siblings.
Perforin- and Granzymes-Producing CD8 T Lymphocytes
The EM lymphocytes, cells enriched in perforin and granzymes,
are reduced in AC and HSAC. As a consequence we expected these
proteins to be diminished in AC and HSAC individuals. Results
confirmed that perforin-producing CD8 T cells were reduced in
AC and HSAC compared with HC children, even if these differences

Figure 3. Cytokine producing CD14 cells in AC,

HSAC, and HC without familiarity for autism. Mitogenstimulated results (background subtracted) are presented. The boxes stretch from the 25th to the 75th
percentile; the lines across the boxes indicate the median values; the lines stretching from the boxes indicate extreme values. Statistical significance is shown.
Abbreviations as in Figure 2.

www.sobp.org/journal

982 BIOL PSYCHIATRY 2009;66:978 984

M. Saresella et al.

Table 2. Lymphocyte Subsets in Peripheral Blood of AC, HSAC, and HC

Lymphocyte Populations
T TOT
B TOT
T ATT
NK
CD4 TOT
CD8 TOT
8RO38
CD4/CD8
CD4 CCR7 45RA
Nave
CD4 CCR7 45RA
Central Memory
CD4 CCR7 45RA
Effector Memory
CD4 CCR7 45RA
Terminally Differentiated
CD8 CCR7 45RA
Nave
CD8 CCR7 45RA
Central Memory
CD8 CCR7 45RA
Effector Memory
CD8 CCR7 45RA
Terminally Differentiated

AC
(n 20)

HSAC
(n 15)

HC
(n 20)

69.6
64.872.9
14
10.717.4
.8
.51.4
10.8
6.416.7
36.9
32.343.4
21
19.824.5
.6
.31.2
1.8
1.52
41.7
32.650.5
19.8
1336.7
24.9
21.230.3
8.4a
4.217
52.6a
40.261
6.2
3.617.8
24.8a
1331.5
16.6
3.924.4

69.5
63.573.3
13.2
9.618.6
.5
.51.2
9.2
5.814.1
38.6
34.344.5
21.5
18.423.1
.3
.13.6
1.8
1.62.1
40
35.149.9
20.7
13.526.2
30.8
18.534.5
5.9b
2.812.5
50b
38.454.1
5.3
3.57.9
29.6b
2535.7
16.3
9.222.9

64.65
59.870.85
16.65
13.819
1.2
.751.65
11.8
7.914.2
31.15
28.938.2
18.4
15.722.5
1.2
.92.95
1.85
1.552.13
34.05
2.745.1
17.5
3.2524.1
25.05
17.9559.75
18.35a,b
13.823.6
25.35a,b
15.2532.75
8.5
2.611.7
42.45a,b
34.95.3551.55
15.1
10.132.55

p .013a
p .002b
p .0001a
p .001b
p .00011a
p .03b

Median and interquartile ranges (% of cells) are shown. All p values are Bonferroni-corrected.
Abbreviations as in Table 1.
AC vs. HC.
b
HSAC vs. HC.
a

did not reach statistical significance; no differences were detected in

granzymes-producing CD8 T cells (Table 1).
Serum Autoantibodies
All sera were antibodies-negative with two exceptions: one
GM1-specific IgG in one AC child, and one GM1-specific IgM in
another AC patient (data not shown). Sera from HSAC and HC
were negative for all the autoantibodies measured (data not
shown).
Virological Analysis
To exclude the presence of infectious diseases at enrollment,
we performed a serologic study for searching IgM antibodies
directed against measles, HSV 1/2, VZV, toxoplasma, parvovirusB19 and EBV-VCA. Parvovirus-B19-specific IgM were observed
in serum of one AC child (data not shown).

Discussion
Subpopulations of T lymphocytes and monocyte-macrophages
that produce proinflammatory cytokines as well as post-thymic
differentiation, are altered in AC compared with HC children. Even
more importantly, a similar immune dysregulation is detected in
healthy, non-autistic siblings of AC children: these results are novel
www.sobp.org/journal

and unexpected. The sharing of common environmental and at

least partially genetic backgrounds between autistic children and
their healthy siblings clearly results in different clinical outcomes: an
autistic condition or lack thereof. It has nevertheless been proposed
that an autism endophenotype characterized by subtle neurologic
impairments can be evidenced in HSAC (2225). Thus, the interaction among genes, environment, and other still undefined factors
can result in full-blown autistic disease or can be associated with
faint, subclinical features. The data herein, although preliminary,
indicate that the autism endophenotype extends its effects to the
immune system.
Cytokines and chemokines modulate brain function, affecting
cognitive and emotional processing (30). Cytokines might be
responsible for features of autism, such as mood and sleep
disturbances. Proinflammatory cytokines, including IL-12 and
IFN- are augmented in plasma of AC children (3133), in whom
TNF cerebrospinal fluid concentrations are also higher (34).
Additionally, increased quantities of TNF and IFN- in blood
and mucosal lymphocytes derived from ileal lymphoid tissue
were described in autistic patients (35). These results are in line
with the hypothesis that neuroinflammation plays a role in the
pathogenesis of autism (36,37). Immune cells can migrate across
the blood-brain barrier (BBB); additionally the BBB is partially

M. Saresella et al.

BIOL PSYCHIATRY 2009;66:978 984 983

Figure 4. Memory/nave T lymphocyte subpopulations in

AC, HSAC, and HC without familiarity for autism. Nave
(CCR7/CD45RA), effector memory (CCR7/CD45RA),
and terminally differentiated (CCR7/CD45RA) T lymphocytes are presented. The boxes stretch from the 25th
to the 75th percentile; the lines across the boxes indicate
the median values; the lines stretching from the boxes
indicate extreme values. Statistical significance is shown.
Abbreviations as in Figure 1.

permeable to cytokines. Thus, immune alterations observed in

peripheral blood could have a direct effect on afferent neurons
and their functions (38). Results of our pilot study support the
hypothesis that cytokine production is altered in autism by
showing that the populations of cytokine-producing immune
cells are significantly different in AC compared with HC and
demonstrate that a similar immune dysregulation is also observed in healthy unaffected siblings of autistic children. On a
speculative side, it would be interesting to evaluate whether
situations that result in time-limited alterations of the immune
response and cytokine production (e.g., infections, seasonal
allergy episodes) would be associated with changes of autismassociated behavioral parameters in AC and HSAC.
A common feature of both AC and HSAC was the increased
production of IL-10 by all the subpopulations of immune cells
examined. Interleukin-10 has strong anti-inflammatory properties: this finding could thus be interpreted as a way in which the
immune system tries to counterbalance the inflammation present
both in autistic patients and in their unaffected siblings. Finally,
we observed the presence of autoantibodies (anti-GM1) in two
autistic children alone. The low number of antibody-positive AC
children could indicate that autoantibodies are technically difficult to detect in serum or rather that this pathogenic mechanism
plays a minor role in autism.
T lymphocyte post-thymic maturation is also impaired, as in
both AC and HSAC individuals, in whom a skewing toward the
earlier, less differentiated lymphocyte subpopulations was evident. Thus, nave T lymphocytes were augmented, and TD
CD4 T cells as well as EM CD8 T cells were diminished in AC
and HSAC compared with HC children. The EM cells derive from
central memory lymphocytes in the presence of antigen and
cytokines (38,39). If the antigenic load is high, EM differentiate
into TD, lythic effector cells (39,40). The situation seen in AC and
HSAC suggests an increased thymic output, resulting in higher
circulating numbers of nave lymphocytes that can normally
differentiate into CM cells, but encounter problems in their
subsequent maturation. On a speculative basis it could be argued
that the lower quantities of EM and TD cells results in a reduced
ability to handle infectious agents and/or vaccines antigens. An

increment of thymic output could be an attempt to bypass the

diminished quantity of circulating lymphocytes with lytic/apoptotic ability. These observations agree with older data indicating
that the activity of NK cells is reduced in autism (19). Overall, a
subtle general impairment of cell-mediated effector mechanisms
seems to be present in autism.
It is hypothesized that in autism a persistently inappropriate
immune response to antigenic stimuli, recently suggested to be
associated with a particular macrosomic endophenotype, (41
44), stems from a particular genetic background and environmental triggers and results in the development of disease. Our
data support the inappropriateness of the immune response in
AC. Nevertheless the observation that similar immunologic abnormalities are detected in AC and in their unaffected siblings
indicates that the immune alterations of autism can only be
considered as permissive toward disease development in the
presence of other, still undefined factors.
Endophenotypes are simple aspects of a diseasepossibly
governed by few susceptibility genesthat are often observed in
unaffected relatives at a higher rate than in the general population. A cognitive endophenotype characterized by a mild reduction in ideational fluency, a nonverbal weakness, and impairments in planning and cognitive flexibility can be observed in
healthy unaffected relatives of children with autism (2225). Our
results extend these findings to an apparently totally different
biological system and suggest that the immune response is
functionally and phenotypically impaired in healthy siblings of
children with autism. The observation that autism is likely
associated with given alleles encoding within the HLA region, a
region of pivotal importance in immunity, offers a speculative
justification to our results and opens novel research avenues.

This study was supported by grants from Ricerca Finalizzata

(Italian Ministry of Health), 2008 Ricerca Corrente (Italian
Ministry of Health), Italian Ministry of Health (ISS 2007; Conv.
Number 526/D25), the Istituto Superiore di Sanita Programma
Nazionale di Ricerca sull AIDS, The EMPRO and AVIP EC WP6
Projects, the nGIN EC WP7 Project, the Japan Health Science
www.sobp.org/journal

984 BIOL PSYCHIATRY 2009;66:978 984

Foundation, Progetto FIRB RETI: Rete Italiana Chimica Farmaceutica CHEM-PROFARMA-NET [RBPR05NWWC], and Fondazione CARIPLO.
The authors reported no biomedical financial interests or
potential conflicts of interest.
1. Amaral DG, Schumann CM, Nordahl CW (2008): Neuroanatomy of autism. Trends Neurosci 31:137145.
2. Ashwood P, Van de Water J (2004): Is autism an autoimmune disease?
Autoimmun Rev 3:557562.
3. Krause I, He XS, Gershwin ME, Shoenfeld Y (2002): Brief report: Immune
factors in autism: A critical review. J Autism Dev Disord 32:337345. Review.
4. Singh VK, Warren R, Averett R, Ghaziuddin M (1997): Circulating autoantibodies to neuronal and glial filament proteins in autism. Pediatr Neurol
17:88 90.
5. Plioplys AV, Greaves A, Kazemi K, Silverman E (1994): Immunoglobulin
reactivity in autism and Retts syndrome. Dev Brain Dysfunct 7:1216.
6. Bashina VM, Kozlova IA, Kliushnik TP, Simashkova NV, Danilovskaia EV,
Gorbachevskaia NL, et al. (1997): [An elevation in the level of autoantibodies to nerve-growth factor in the blood serum of schizophrenic
children]. Zh Nevropatol Psikhiatr Im S S Korsakova 97:4751.
7. Todd RD, Ciaranello RD (1985): Demonstration of inter- and intraspecies
differences in serotonin binding sites by antibodies from an autistic
child. Proc Natl Acad Sci U S A 82:612 616.
8. Yuwiler A, Shih JC, Chen CH, Ritvo ER, Hanna G, Ellison GW, et al. (1992):
Hyperserotoninemia and antiserotonin antibodies in autism and other
disorders. J Autism Dev Disord 22:33 45.
9. Ferrari P, Marescot MR, Moulias R, Bursztejn C, Deville Chabrolle A,
Thiollet M, et al. (1988): Immune status in infantile autism. Correlation
between the immune status, autistic symptoms and levels of serotonin.
Encephale 14:339 344.
10. Croonenberghs J, Wauters A, Devreese K, Verkerk R, Scharpe S, Bosmans
E, et al. (2002): Increased serum albumin, gamma globulin, immunoglobulin IgG, and IgG2 and IgG4 in autism. Psychol Med 32:14571463.
11. Stubbs EG, Crawford ML, Burger DR, Vandenbark AA (1977): Depressed
lymphocyte responsiveness in autistic children. J Autism Child Schizophr
7:49 55.
12. Plioplys AV, Greaves A, Kazemi K, Silverman E (1994): Lymphocyte function in autism and Retts syndrome. Neuropsicobiology 29:1216.
13. Warren RP, Foster A, Mararetten NV, Pace NC (1986): Immune abnormalities in patients with autism. J Autism Dev Disord 16:189 197.
14. Warren RP, Yonk LJ, Burger RA, Cole P, Odell JD, Warren WL, et al. (1990):
Deficiency of suppressor-inducer (CD4 CD45RA) T cells in autism.
Immunol Investig 19:245251.
15. Yonk LJ, Warren RP, Burger RA, Cole P, Odell JD, Warren WL, et al. (1990):
CD4 helper T cell depression in autism. Immunol Lett 25:341345.
16. Menage P, Thibault G, Barthelemy C, Lelord G, Bardos P (1992): CD4
CD45RA T lymphocyte deficiency in autistic children: Effect of a pyridoxine-Magnesium treatment. Brain Dysfunct 5:326 333.
17. Denney DR, Wood Frei B, Gaffney GR (1996): Lymphocyte subsets and
interleukin-2 receptors in autistic children. J Autism Dev Disord 26:8797.
18. Ferrante P, Saresella M, Guerini FR, Marzorati M, Musetti MC, Cazzullo CL
(2003): Significant association of HLA A2-DR11 with CD4 naive decrease
in autistic children. Biomed Pharmacother 57:372374.
19. Warren RP, Foster A, Mararetten NC (1987): Reduced natural killer cell
activity in autism. J Am Acad Child Adolesc Psychiatry 26:333335.
20. Gupta S, Aggarwal S, Rashanravan B, Lee T (1998): Th1 and Th2-like
cytokines in CD4 and CD8 T cells in autism. J Neuroimmunol 85:106
109.
21. Engstrom AH, Ohlson S, Stubbs EG, Maciulus A, Caldwell V, Odell JD, et
al. (2003): Decreased expression of CD95 (FAS/APO-1) on CD4 T-lymphocytes from participants with autism. J Dev Phys Disabil 15:155163.
22. Constantino JN, Lajonchere C, Lutz M, Gray T, Abbacchi A, McKenna K, et
al. (2006): Autistic social impairment in the siblings of children with
pervasive developmental disorders. Am J Psychiatry 163:294 296.

www.sobp.org/journal

M. Saresella et al.
23. Constantino JN, Todd RD (2005): Intergenerational transmission of subthreshold autistic traits in the general population. Biol Psychiatry 57:
655 660.
24. Alarcn M, Yonan AL, Gilliam TC, Cantor RM, Geschwind DH (2005):
Quantitative genome scan and ordered-subsets analysis of autism endophenotypes support language QTLs. Mol Psychiatry 10:747757.
25. Yirmiya N, Ozonoff S (2007): The very early autism phenotype. J Aut Dev
Disord 37:111.
26. Torres AR, Maciulis A, Stubbs EG, Cutler A, Odell D (2002): The transmission disequilibrium test suggests that HLA-DR4 and DR13 are linked to
autism spectrum disorder. Hum Immunol 63:311322.
27. Lee LC, Zachary AA, Leffell MS, Newschaffer CJ, Matteson KJ, Tyler JD, et
al. (2006): HLA-DR4 in families with autism. Pediatr Neurol 35:303316.
28. Guerini FR, Manca S, Sotgiu S, Tremolada S, Zanzottera M, Agliardi C, et
al. (2006): A family based linkage analysis of HLA and 5-HTTLPR gene
polymorphisms in Sardinian children with autism spectrum disorder.
Hum Immunol 67:108 120.
29. Nemni R, Fazio R, Quattrini A, Lorenzetti I, Mamoli D, Canal N (1993):
Antibodies to sulfatide and to chondroitin sulfate C in patients with
chronic sensory neuropathy. J Neuroimmunol 43:79 86.
30. Ashwood P, Wills S, Van de Water J (2006): The immune response in
autism: A new frontier for autism research. J Leukoc Biol 80:115.
31. Singh VK (1996): Plasma increase of interleukin-12 and interferongamma. Pathologic significance in autism. J Neuroimmunol 66:143145.
32. Jyonouchi H, Sun S, Le H (2001): Proinflammatory and regulatory cytokine production associated with innate and adaptive immune responses in children with autism spectrum disorders and developmental
regression. J Neuroimmunol 120:170 179.
33. Croonenberghs J, Bosmans E, Deboutte D, Kenis G, Maes M (2002):
Activation of the inflammatory response system in autism. Nueropsycobiology 45:1 6.
34. Chez MG, Dowling T, Patel PB, Khanna P, Kominsky M (2007): Elevation
of tumor necrosis factor-alpha in cerebrospinal fluid of autistic children.
Pediatr Neurol 36:361365.
35. Ashwood P, Wakefield AJ (2006): Immune activation of peripheral blood
and mucosal CD3 lymphocyte cytokine profiles in children with autism and gastrointestinal symptoms. J Neuroimmunol 173:126 134.
36. Snider LA, Swedo SE (2003): Post-streptococcal autoimmune disorders
of the central nervous system. Curr Opin Neurol 16:359 365.
37. Morris CM, Pardo-Villamizar C, Gause CD, Singer HS (2009): Serum autoantibodies measured by immunofluorescence confirm a failure to differentiate PANDAS and Tourette syndrome from controls. J Neurol Sci
276:45 48.
38. Dantzer R, Kelley KW (2007): Twenty years of research on cytokineinduced sickness behavior. Brain Behav Immun 21:153160.
39. Sallusto F, Langenkamp A, Geginat J, Lanzavecchia A (2000): Functional
subsets of memory T cells identified by CCR7 expression. Curr Top Microbiol Immunol 251:167171.
40. Geginat J, Sallusto F, Lanzavecchia A (2001): Cytokine-driven proliferation and differentiation of human naive, central memory, and effector
memory CD4() T cells. J Exp Med 194:17111719.
41. Sacco R, Militerni R, Frolli A, Bravaccio C, Gritti A, Elia M, et al. (2007):
Clinical, morphological, and biochemical correlates of head circumference in autism. Biol Psychiatry 62:1038 1047.
42. Beasley R, Crane J (2001): Head circumference at birth and the development of allergic diseasemore questions than answers. Clin Exp Allerg
31:1804 1805.
43. Katz KA, Pocock AJ, Strachan DP (2003): Neonatal head circumference,
neonatal weight, and risk of hayfever, asthma and eczema in a large cohort
of adolescents from Sheffield, England. Clin Exp Allerg 33:737745.
44. Bernsen RM, de Jongste JC, Koes BW, Aardoom HA, van der Wouden JC
(2005): Perinatal characteristics and obstetric complications as risk factors for asthma, allergy and eczema at the age of 6 years. Clin Exp Allerg
35:11351140.