I.
Introduction
Various
cytokines have been known to form the network between the immune system and the
central nervous system. These cytokines show a change not only in the
peripheral immune system but also in the central nervous system, and they exert
their effects on the proliferation and death of immune cells; they are
associated with neurotransmitters metabolism, neuronal development and
degeneration (Muller and Ackenheil, 1998). TNF-α is a very important
pro-inflammatory cytokine involved in the regulation of immune system and in
the initial stage of inflammatory reaction, and it has been related to abnormal
behaviors, anxiety, decreased appetite, emotions, sleep disturbances,
recognition and other psychiatric symptoms highly expressed also in
schizophrenia (Holden and Pakula, 1999;Kubota et al., 2001;Reichenberg et al.,
2001;Strieter et al., 1993). In addition, the serum concentration of
TNF-α in schizophrenic patients has been reported to be elevated in
comparison with the control group. Moreover, after the administration of
antipsychotic drugs, an alteration of its concentration has been detected, and
thus its involvement not only in the development of schizophrenia but also in
the response to treatment has been suggested (Erbagci et al., 2001;Monteleone et al., 1997;Naudin et al., 1997). TNF-α gene (TNFA) is located on
chromosome 6p21.3 and was reported to be associated with schizophrenia (Schwab et al., 2000); the TNFA –238 (G/A) and
–308 (G/A) polymorphisms are involved in the regulation of TNF-α
activation, by influencing TNF-α transcription, thus being involved
directly in the production of TNF-α (Kaluza et al., 2000;Wilson et al., 1994). According to the association of the
TNFA polymorphism and schizophrenia, a correlation has been reported in
literatures, with conflicting results (Boin et al., 2001;Czerski et al., 2008;Duan et al., 2004;Handoko et
al., 2003;Hanninen et al., 2005;Hashimoto et al., 2004;Kampman et al.,
2005;Meira-Lima et al., 2003;Morar et al., 2007;Pae et al., 2003;Riedel et al.,
2002;Saviouk et al., 2005;Schwab et al., 2003;Shirts et al., 2006;Tan et al.,
2003;Tsai et al., 2003;Watanabe et al., 2007;Zai et al., 2006), although a recent meta-analyses including 2512 cases and 3223 controls
showed that the AA genotype was weakly associated with schizophrenia
susceptibility in Caucasoids (Odd Ratio OR=1.65, 95% CI=1.00-2.71 Z=1.98
p=0.05) (Sacchetti et al., 2007).
To our
knowledge, however, investigations about the association between TNF-α
polymorphism and antipsychotic treatment response have been still inadequate to
date. The aim of this study is to investigate the presence of an interaction
effect of the TNF-α –308 and -238 polymorphisms on the development
of schizophrenia and on the clinical symptomatology. Firstly we investigated
the association between individual polymorphisms and the vulnerability to
schizophrenia and to diverse clinical variables. Then, we compared and analyzed
the influence of the interaction effects of the two polymorphisms on the
affection status and various clinical variables.
II.
Methods
A. Subjects
One hundred and fifty-two (152)
inpatients with schizophrenia and the 152 healthy controls participated in this
study. The diagnosis was made on the basis of consensus between two
board-certified psychiatrists according to the Diagnostic and Statistical
Manual of Mental Disorders-IV (DSM-IV) (American,
1994). The Structured
Clinical Interview for DSM-IV Axis I Disorders-Clinician Version (SCID-I-CV) (First
et al., 1997) was
administered to all patients. Other available resources such as clinical
course, family information and medical records were also used. Patients with a
diagnosis of other mental or neurological disorders other than schizophrenia
were excluded from the study. The control group consisted of volunteers from
the paramedical- and medical-staffs or students who showed no family or
personal history of major mental disorders and neurological diseases. The
control group was also evaluated by board certified-psychiatrists to exclude
current or a past history of psychiatric problems. A written informed-consent
form was obtained from all subjects after a full description of the study. The
Positive and Negative Syndrome Scale (PANSS) (Kay
et al., 1988) was assessed
for examining the antipsychotics treatment response in patients with
schizophrenia at the time of admission and 8 6weeks after antipsychotic
administration (responder: defined as 20% decrease of PANSS score after 8-week
treatment compared to that of PANSS score at the time of admission). Any
antipsychotics regardless of class were given for the treatment of psychotic
symptoms for 8 weeks. Risperidone (n=71) was the most frequently administered
antipsychotic and then olanzapine (n=39), quetiapine (17), chlorpromazine (13)
and haloperidol (12) were followed. One hundred and thirty-one (86.2%) patients
completed the 8-week treatment. Thus, the 21 patientsÕ last data was obtained
by a method of the last observation carried forward (LOCF).
B. Genotyping with Restriction Fragment Length Polymorphism
(RFLP)
The DNA was extracted from whole blood
using the standard method. Amplification of the target site of the TNF-α
gene (TNFA –238, rs 361525 and –308, rs1800629) was carried out by
PCR using forward (F) and reverse (R) primers (TNFA –238: F5'-ATC TGG AGG
AAG CGG TAG TG-3' and R5'-AGA AGA CCC CCC TCG GAA CC-3'; TNFA –308:
F5'-AGGCAATAGGTTTTGAGGGCCAT-3' and R5'-TCCTCCCTGCTCCGATTCCG-3') and conditions
as described elsewhere (Wilson
et al., 1994). The PCR
products were digested at 37¡C with NcoI
(Boehringer Mannheim, Mannheim, Germany)
to detect the TNFA –308 allele and with MspI (Boehringer Mannheim, Mannheim, Germany)
to detect the TNFA –238 allele and then subjected to 8% acrylamide gel
and stained with ethidium bromide. As for TNFA –238, Large 7(152bp) and
small (132 bp/20 bp) fragments were identified as MspI site negative (TNFA –238 A allele) and positive (TNFA
–238 G allele), respectively. As for TNFA –308, Large (107bp) and
small (87 bp/20 bp) fragments were identified as NcoΙ site negative (TNFA –308 A allele) and positive
(TNFA –308 G allele), respectively.
C. Statistical Analyses
Differences in the allele and genotype
frequencies of the TNFA-238 and -308 polymorphisms between the patients and the
controls were calculated using a Chi-square test. Hardy-Weinberg equilibrium at
each polymorphism was examined by a goodness-of-fit Chi-square. To compare
numerical variable such as age and the number of admission, nonparametric
t-test, Univariate Analysis of Variance (ANOVA) and ANCOVA were used where
appropriate. The Monte Carlo method with the CLUMP program v 1.9 (10,000
simulations) (Sham
and Curtis, 1995) was used when
small cell count observed in case of categorical comparison. Multiple
regression analysis was applied to detect the possible interaction effects
between the TNFA –238 and -308 polymorphisms in their influences on the
affection status and other clinical variables (Serretti
et al., 1999). The Linkage
Disequilibrium (LD) was tested using a two-locus LD calculator (2 LD) (Zhao,
2004). The haplotypes
were constructed using the EH program (Xie
and Ott, 1993). P values
<0.05 were considered significant. All the statistical tests were performed
using the SPSS v10.0 software (SPSS Inc., Chicago, IL). The power of a sample
to detect the differences between the alleles was calculated by considering an
alpha value of 0.05, two tailed. Using these parameters, the sample was
estimated to have the power (0.80) to detect a small to medium effect size
(w=0.16), which corresponded to a difference of approximately 15% between the
two groups in the two alleles.
III. Results
Seventy-three
(48.0%) patients with schizophrenia were male and 69 (45.4%) normal controls
were male. No difference was present in the gender distribution between the two
groups (p=0.730). The average age was older in the patient group (mean ± SD,
36.9±11.6 years) than in the control group (32.0±10.7 years) (p < 0.05).
The
frequencies of TNFA -238 and -308 genotypes were not different from the
expected values of Hardy-Weinberg equilibrium in the patients group (p=0.170
and p=0.883, respectively) and in the control group (p=0.112 and p=0.091,
respectively, Table 1). Neither TNFA
-238 nor TNFA -308 polymorphism variants were associated with the development
of schizophrenia (Table 1).
The
description of the patientsÕ characteristics with schizophrenia according to
polymorphisms of TNFA -238 and TNFA -308 is shown in Tables 2 and 3. Subjects
carrying TNFA -238 A allele had more frequent familial history [genotype AA and
AG, p=0.023, odds ratio (OR)=5.156, 95% confidence intervals (CI)=1.463-18.107]
(Table 2). Neither TNFA –238
polymorphism nor TNFA -308 variants were associated with the antipsychotic
treatment response as a categorical variable (Tables 2 and 3).
Other
clinical variables such as onset age, suicide history, past admission number
and duration of illness were not associated with the two polymorphisms. When
analyzing with the multiple regression approach all possible interactions
between all the possible combinations of the two polymorphisms, significant
interaction effects were not found with the development of schizophrenia as
well as with several clinical variables such as family history, attempted
suicide history, age at onset, duration of illness and antipsychotic treatment
response. However, interaction effect between TNFA -238 AG genotype and
–308 AA genotype was found to be associated with family history (beta=-0.432,
p=0.017).
The D′ coefficient for the TNFA -238 and -308
polymorphisms was not indicative of significant LD (D′ =0.564) so that the haplotype analysis
was not presented.
Table
1: The distribution of
genotypeand allele in patients with schizophrenia and the controls
Table 2: Clinical parameters in patients with
schizophrenia
according to TNFA -238 polymorphism
Table 3: Clinical parameters in patients with
schizophrenia
according to TNFA -308 polymorphism
IV.
Discussion
As for
the association of TNFA polymorphism with schizophrenia, there have been mixed
results, some studies supported (mainly with TNFA –308 polymorphism) (Czerski et al., 2008;Meira-Lima et al., 2003;Sacchetti et al.,
2007;Saviouk et al., 2005;Schwab et al., 2000;Tan et al., 2003), while others failed to replicate the
association of TNFA polymorphism with schizophrenia (Duan et al., 2004;Handoko
et al., 2003;Hashimoto et al., 2004;Pae et al., 2003;Riedel et al., 2002;Shirts
et al., 2006;Tsai et al., 2003;Watanabe et al., 2007;Zai et al., 2006). Since Boin et al. (Boin et al., 2001) found that TNFA –308 A allele was
highly associated with a susceptibility to schizophrenia, which were replicated
in other studies with different ethnicities (Caucasian and Brazilian and only
one for Chinese) (Meira-Lima et al., 2003;Schwab et al., 2000;Tan et al., 2003).
Among
the positive studies, interestingly, some found the positive association with
TNFA -308 A allele with the susceptibility to schizophrenia, while others
reported the positive association with TNFA -308 G allele, suggesting a
complicated role of TNFA –308 polymorphism for schizophrenia.
Furthermore, a pooled analysis with 2,399 patients with schizophrenia and 3,261
controls found that the TNFA –308 polymorphism was not associated with
the development of schizophrenia, same result was given when the sample was
stratified into different ethnics by the geographical origin though (i.e, Asian
vs. Caucasian), which implicate the ethnic heterogeneity of the polymorphism (Duan et al., 2004).
Taken
former studies together, our present findings regarding the TNFA –238 and
-308 polymorphisms would be in the line with the majorities that did not prove
the association of the TNFA -238 and -308 polymorphisms with schizophrenia (Boin et al., 2001;Duan et al., 2004;Handoko et al., 2003;Hanninen
et al., 2005;Hashimoto et al., 2004;Meira-Lima et al., 2003;Pae et al.,
2003;Riedel et al., 2002;Schwab et al., 2003;Tan et al., 2003;Tsai et al., 2003). Considering the lower frequency of A
alleles in the two polymorphisms, we may need to collect larger samples and to
conduct family-based studies. It should be also noted that the LD between TNFA
–238 and -308 polymorphisms were not found in the present study, which is
in line with the previous study (Duan et al., 2004), indicating less information of
haplotype construction of the two polymorphisms for the development of
schizophrenia until now, although available information on the LD on the TNFA
polymorphisms would enhance the knowledge about the possible causative genes (Duan et al., 2004;Handoko et al., 2003;Tan et al., 2003). We found the subjects carrying TNFA
-238 A allele (genotype AA or AG) were associated with positive familial
history, suggesting this polymorphism would be an indicator for familial
tendency of schizophrenia. However, the sub-analyses were performed with very
small samples so that we need to be cautious in the interpretation of this
finding. There have been little studies in relation to TNFA polymorphisms and
antipsychotic treatment response. Recent study (Tsai et al., 2003) reported that TNFA –308
polymorphism might not be associated with clozapine response. Our study also
failed to find any significant association of the two polymorphisms and
antipsychotic response. This may suggest the TNFA -238 and -308 polymorphisms
would not be associated with the improvement of psychopathology, although, we
need more pharmacogenetic data with TNFA polymorphisms. In fact a recent study
showed a marginal association of TNFA with antipsychotic response (Zai et al., 2006), indicating a need of more data in this
area.
We tried
to find possible interactions effects, unfortunately, we failed to find
significant interactions except for positive association of the TNFA -238 AG
and –308 AA with positive family history. As previously stated, the
subjects with TNFA -238 allele A (genotype AA and AG) were associated with
positive familial history, which may indicate the strong effect of TNFA
–238 polymorphism on the familial aggregation for the development of schizophrenia
considering the result of interaction and individual effects. Other interaction
effects of the two polymorphisms on the other clinical variables were not found
to be associated in patients with schizophrenia. The interaction between the
two polymorphisms may not play a major role in development of schizophrenia and
major clinical manifestations, or has only weak genetic influence. Different
ethnic background for the two polymorphisms should be considered (Boin et al., 2001;Duan et al., 2004;Handoko et al., 2003;Meira-Lima
et al., 2003;Pae et al., 2003;Tsai et al., 2003). Case-control association study has
inherent pitfall of population stratification. More SNPs from the two genes
should provide more accurate information. The multiple comparison issue should
also be noted because the two polymorphisms were compared with regard to the
genotype frequency and the allele distribution as well as the correlation with
clinical variables including numerical and categorical parameters. We should
consider the small sample number considering conventional sample size needed in
the case-control association study (Sham and Curtis, 1995) which is related to the possibility of
false negative finding with regard to sample size. Finally this study has only investigated
the possibility of immunogenetical perspective in relation to pathophysiologies
of schizophrenia but not tried to find other pathophysiological aspects such as
neurotransmitters or neuroendocrine factors. Although, interaction effects may
also partly reveal the potential role of specific genes for certain
multi-factorial disorders, epistatic interaction among candidate genes for
certain mental disorders might have us more chance to understand their genetic
backgrounds and molecular underpinnings than simple interaction study.
In
conclusion, the present study suggest that the interaction effects between TNFA
-238 and -308 polymorphisms gives no significant contribution to the
susceptibility to schizophrenia, and are not associated with clinical variables
such as antipsychotic treatment response, except for family history, at least
in Korean population. Consecutive studies with larger sample would be needed to
draw any confirmative conclusion.
Acknowledgment
This study was supported by a grant of the Korean
Health 21 R & D Project, Ministry of Health and Welfare, Republic of Korea
(03-PJ10-PG13-GD01-0002 or A030001) and a support from the Medical Research
Center, Korea Science and Engineering Foundation, Republic of Korea (R13-2002-005-04001-0).
References
American PA
(1994). "Diagnostic and Statistical Manual of Mental Disorder, fourth
ed.,." American Psychiatric Press, Washington DC.
Boin F, Zanardini R, Pioli
R, Altamura CA, Maes M, and Gennarelli M (2001)
Association between -G308A tumor necrosis factor alpha gene polymorphism and
schizophrenia. Mol Psychiatry 6,
79-82.
Czerski PM, Rybakowski F,
Kapelski P, Rybakowski JK, Dmitrzak-Weglarz M, Leszczynska-Rodziewicz A,
Slopien A, Skibinska M, Kaczmarkiewicz-Fass M, and Hauser J (2008) Association of tumor necrosis
factor -308G/A promoter polymorphism with schizophrenia and bipolar affective
disorder in a Polish population. Neuropsychobiology
57, 88-94.
Duan S, Xu Y, Chen W, Liu Z,
Guo T, Gao J, Bian L, Zhen Y, Li X, Zhang X, Pan Y, Gu N, Feng G, and He L (2004) No association between the
promoter variants of tumor necrosis factor alpha (TNF-alpha) and schizophrenia
in Chinese Han population. Neurosci Lett
366, 139-43.
Erbagci AB, Herken H,
Koyluoglu O, Yilmaz N, and Tarakcioglu M (2001)
Serum IL-1beta, sIL-2R, IL-6, IL-8 and TNF-alpha in schizophrenic patients,
relation with symptomatology and responsiveness to risperidone treatment. Mediators Inflamm 10, 109-15.
First MB, Spitzer RL, Gibbon
M, and Williams JBW (1997). "Structured Clinical Interview for
DSM-IV-Clinician Version (SCID-CV)." American Psychiatric Press,
Washington DC.
Handoko HY, Nancarrow DJ,
Hayward NK, Ohaeri JU, Aghanwa H, McGrath JJ, Levinson DF, Johns C, Walters MK,
Nertney DA, Srinivasan TN, Thara R, and Mowry BJ (2003) Tumor necrosis factor haplotype analysis amongst
schizophrenia probands from four distinct populations in the Asia-Pacific
region. Am J Med Genet B Neuropsychiatr
Genet 121B, 1-6.
Hanninen K, Katila H, Rontu
R, Mattila KM, Hurme M, and Lehtimaki T (2005)
Tumor necrosis factor-alpha --G308A polymorphism in schizophrenia in a Finnish
population. Neurosci Lett 385,
76-81.
Hashimoto R, Yoshida M,
Ozaki N, Yamanouchi Y, Iwata N, Suzuki T, Kitajima T, Tatsumi M, Kamijima K,
and Kunugi H (2004) Association
analysis of the -308G>A promoter polymorphism of the tumor necrosis factor
alpha (TNF-alpha) gene in Japanese patients with schizophrenia. J Neural Transm 111, 217-21.
Holden RJ, and Pakula IS (1999) Tumor necrosis factor-alpha: is
there a continuum of liability between stress, anxiety states and anorexia
nervosa? Med Hypotheses 52, 155-62.
Kaluza W, Reuss E, Grossmann
S, Hug R, Schopf RE, Galle PR, Maerker-Hermann E, and Hoehler T (2000) Different transcriptional
activity and in vitro TNF-alpha production in psoriasis patients carrying the
TNF-alpha 238A promoter polymorphism. J
Invest Dermatol 114, 1180-3.
Kampman O, Anttila S, Illi
A, Mattila KM, Rontu R, Leinonen E, and Lehtimaki T (2005) Interaction of tumor necrosis alpha - G308A and epidermal
growth factor gene polymorphisms in early-onset schizophrenia. Eur Arch Psychiatry Clin Neurosci 255,
279-83.
Kay SR, Opler LA, and
Lindenmayer JP (1988) Reliability
and validity of the positive and negative syndrome scale for schizophrenics. Psychiatry Res 23, 99-110.
Kubota T, Fang J, Guan Z,
Brown RA, and Krueger JM (2001)
Vagotomy attenuates tumor necrosis factor-alpha-induced sleep and EEG
delta-activity in rats. Am J Physiol
Regul Integr Comp Physiol 280, R1213-20.
Meira-Lima IV, Pereira AC, Mota
GF, Floriano M, Araujo F, Mansur AJ, Krieger JE, and Vallada H (2003) Analysis of a polymorphism in the
promoter region of the tumor necrosis factor alpha gene in schizophrenia and
bipolar disorder: further support for an association with schizophrenia. Mol Psychiatry 8, 718-20.
Monteleone P, Fabrazzo M,
Tortorella A, and Maj M (1997)
Plasma levels of interleukin-6 and tumor necrosis factor alpha in chronic
schizophrenia: effects of clozapine treatment. Psychiatry Res 71, 11-7.
Morar B, Schwab SG, Albus M,
Maier W, Lerer B, and Wildenauer DB (2007)
Evaluation of association of SNPs in the TNF alpha gene region with
schizophrenia. Am J Med Genet B
Neuropsychiatr Genet 144B, 318-24.
Muller N, and Ackenheil M (1998) Psychoneuroimmunology and the
cytokine action in the CNS: implications for psychiatric disorders. Prog Neuropsychopharmacol Biol Psychiatry
22, 1-33.
Naudin J, Capo C, Giusano B,
Mege JL, and Azorin JM (1997) A
differential role for interleukin-6 and tumor necrosis factor-alpha in
schizophrenia? Schizophr Res 26,
227-33.
Pae CU, Chae JH, Bahk WM,
Han H, Jun TY, Kim KS, Kwon YS, and Serretti A (2003) Tumor necrosis factor-alpha gene polymorphism at position
-308 and schizophrenia in the Korean population. Psychiatry Clin Neurosci 57, 399-403.
Reichenberg A, Yirmiya R,
Schuld A, Kraus T, Haack M, Morag A, and Pollmacher T (2001) Cytokine-associated emotional and cognitive disturbances in
humans. Arch Gen Psychiatry 58,
445-52.
Riedel M, Kronig H, Schwarz
MJ, Engel RR, Kuhn KU, Sikorski C, Sokullu S, Ackenheil M, Moller HJ, and
Muller N (2002) No association
between the G308A polymorphism of the tumor necrosis factor-alpha gene and
schizophrenia. Eur Arch Psychiatry Clin
Neurosci 252, 232-4.
Sacchetti E,
Bocchio-Chiavetto L, Valsecchi P, Scassellati C, Pasqualetti P, Bonvicini C,
Corsini P, Rossi G, Cesana BM, Barlati S, and Gennarelli M (2007) -G308A tumor necrosis factor
alpha functional polymorphism and schizophrenia risk: meta-analysis plus
association study. Brain Behav Immun
21, 450-7.
Saviouk V, Chow EW, Bassett
AS, and Brzustowicz LM (2005) Tumor
necrosis factor promoter haplotype associated with