deletion polymorphisms in the nogo gene and chronic schizophrenia

Molecular Brain Research 139 (2005) 212 – 216 www.elsevier.com/locate/molbrainres

Research Report

Gender-specific association of insertion/deletion polymorphisms in the nogo gene and chronic schizophrenia Ene-Choo Tana,d,*, Siow-Ann Chongb, Hanhui Wangb, Eileen Chew-Ping Lima, Yik-Ying Teoc a

Defence Medical and Environmental Research Institute, DSO National Laboratories, 27 Medical Drive, Kent Ridge, 117510, Singapore b Department of Early Psychosis Intervention, Woodbridge Hospital/Institute of Mental Health, Singapore c Department of Statistics, University of Oxford, UK d Department of Psychological Medicine, National University of Singapore, Singapore Accepted 12 May 2005 Available online 13 June 2005

Abstract Nogo is a myelin-associated protein associated with neurite outgrowth and regeneration. A previous study has reported an association between an insertion/deletion polymorphism in schizophrenia. We tested for the distribution of the polymorphism and haplotypes of this and another insertion/deletion polymorphism in our population. We have also developed an assay combining allele-specific polymerase chain reaction (AS-PCR) and restriction fragment length polymorphism (RFLP) to simultaneously type these two insertion/deletion polymorphisms. There was a statistically significant difference at the allelic level for both the CAA (v 2 = 4.378, df = 1, P value = 0.036) and TATC (v 2 = 5.807, df = 1, P = 0.016) polymorphisms in the female subgroup, but not in males. With our genotyping method, we also determined the molecular haplotype. Within the female gender, odds ratio is at 1.57 (95% CI 1.05 – 2.37) for CAACAA-TATC and 1.40 (95% CI 0.55 – 3.60) for CAA-TATC, the two at-risk haplotypes. Odds ratio is 0.63 (95% CI 0.42 – 0.93) for the protective wildtype haplotype CAA-TATCTATC. Further study of these two polymorphisms to investigate functional significance and confirm gender-specific association should be carried out. D 2005 Elsevier B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Neuropsychiatric disorders Keywords: Schizophrenia; Genetic polymorphism; Nogo; Chinese; Female

1. Introduction It is well established that axons of the CNS of higher vertebrates such as man are limited in their capacity to regenerate and undergo repair after injury or assault. This lack of regeneration is associated with a number of proteins which inhibit the growth of the axons. These proteins include myelin-associated glycoprotein (MAG), reticulon 4 (RTN4 or Nogo), and oligodendrocyte-myelin glycoprotein (OMgp). Together with other growth and transcription factors, they regulate the regenerative and repair processes

* Corresponding author. Fax: +65 64857262. E-mail address: [email protected] (E.-C. Tan). 0169-328X/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.molbrainres.2005.05.010

in the central nervous system by inhibiting the outgrowth of neurites and growth cones. Nogo was first identified as a high molecular weight myelin inhibitor. The nogo gene with 14 exons spanning 75 kb is located on chromosome 2p13– 14 [22]. It encodes three major transcripts corresponding to 3 protein variants, Nogo-A (1192 amino acid residues), Nogo-B (373 aa), and Nogo-C (199 aa) [2]. Additionally, there might be as many as 10 splice variants whose mRNAs have been documented although the existence of the proteins are still not clear [16]. The CNS-specific Nogo-A has a unique amino-terminal domain, while the 66-aa extracellular domain and the carboxyl-terminal are shared by all 3 variants. All three bind to a common Nogo receptor (NgR). Nogo-A is expressed at high levels in both oligodendrocytes and

E.-C. Tan et al. / Molecular Brain Research 139 (2005) 212 – 216

neurons [10]. IgM autoantibodies to the recombinant large N-terminal inhibitory domain of Nogo-A have been reported in patients with multiple sclerosis, acute inflammatory, and non-inflammatory neurological diseases [17]. The levels and distribution of Nogo mRNA have been shown to be altered in the cortex of patients with schizophrenia [15], making it an excellent candidate gene for the disorder as deficiencies in cortical functions are well documented in patients. In addition to functional significance, the nogo gene is also a positional candidate, being located within the region 2p12 – 15 for which linkage to schizophrenia families have been reported [1,4,13,19]. Novak et al. also reported an association between schizophrenia and the homozygous CAA insertion genotype in a small set of mostly Caucasoid samples [15]. In order to follow up on their observation, we investigated the frequencies of this triplet insertion and an additional tetranucleotide insertion/deletion polymorphism in the nogo gene in a larger set of controls and patients with schizophrenia. We have also developed simple PCR-RFLP and ARMS-PCR assays for molecular haplotype information.

2. Materials and methods 2.1. Subjects Unrelated patients with schizophrenia (n = 363) were recruited from the in-patient wards for the study (270 males and 93 females with mean age 53.3 T 10.5 years). They were assessed independently by two psychiatrists to have met the DSM-IV criteria for schizophrenia. Controls numbered 253 (124 males and 129 females aged 39.2 T 16.9 years) were volunteers from hospital and institute research staff with no history of psychiatric disorders. All were ethnic Chinese which was defined as having both parents and all four grandparents who were Chinese. Participants gave informed consent and the study was approved by the Hospital Ethics Committee. 2.2. PCR amplification and restriction endonuclease analysis Venous blood was collected in EDTA tubes and genomic DNA extracted using the QIAamp Blood Kit (Qiagen GmbH, Hilden, Germany). Target DNA was then amplified with two outer primers (5V-TTACCTGTCTTGACTGCC-3V and 5V-TACAGCTTAAACCACAATGG-3V according to Novak et al.) and two allele-specific inner primers (5VTGAAATTGATGTTGTTGGA-3Vfor CAA insertion and 5VCACACATAGAACTCCAACAT-3Vfor no CAA insertion). Polymerase chain reaction was carried out in a volume of 10 Al consisting of genomic DNA, 5% glycerol, 2 mM of each of the 4 deoxyribonucleotides, 1 pmol of each of the two outer primers and 10 pmol of each of the two inner primers, 10 mM Tris –HCl pH 8.3, 50 mM KCl, 1.5 mM MgCl2, and

213

1 unit of Taq polymerase (Promega Corporation, Madison, WI, USA). Amplicon sizes expected are 738 bp from the outer primers, 584 bp if the CAA insertion is present or 185 bp if there is no CAA insertion. For genotyping of the TATC insertion/deletion polymorphism, the PCR products were incubated with the restriction enzyme EcoNI according to the manufacturer’s recommendations (New England BioLabs, Beverly, MA, USA). Product sizes would be 649 bp, 495 bp, and 89 bp if the allele with only one copy of TATC is present. For the other allele TATCTATC, the fragments would not be cleaved and the products remain at 738 bp and 584 bp. (In both cases, the 158-bp amplified product from the wildtype CAA allele might also be present.) Amplified and restricted products were separated by agarose gel electrophoresis and photographed under ultraviolet light with ethidium bromide staining. Alleles and genotypes were scored manually by gene counting. 2.3. Statistical analysis Chi-square analysis and Fisher’s exact test were used for comparisons between groups and testing for conformation to Hardy– Weinberg Equilibrium. Odds ratio (OR) and respective 95% confidence intervals were reported to evaluate the effects of any difference between alleles, genotypes, and haplotypes. Probability values of 0.05 or less were regarded as statistically significant. In addition to molecular haplotyping, the 2 polymorphisms were also genotyped and scored separately. Frequencies of haplotypes were then inferred using the Expectation – Maximization (EM) algorithm. Statistical significance for the strength of linkage disequilibrium was obtained empirically through Monte Carlo permutation tests of 10,000 iterations. All statistical analysis was performed on the statistical software S-Plus Professional.

3. Results Allele and genotype frequencies for the two polymorphisms for 363 patients with schizophrenia and 253 healthy controls are shown in Table 1. The observed genotype distribution for both polymorphisms did not deviate from those expected under Hardy –Weinberg equilibrium (CAAcontrols: v 2 = 0.787, P = 0.374, CAA-cases: v 2 = 0.047, P = 0.828, TATC-controls: v 2 = 1.729, P = 0.188, TATC-cases: v 2 = 1.680, P = 0.999). In accordance with the finding of Novak et al., the CAA insertion was found to be more common in the patients although the difference did not reach statistical significance. As there was significant difference in gender distribution between the control and patient groups, analysis was performed for males and females separately. There was no positive association for males. For females, there was a statistically significant association for both polymorphisms at the allelic level (CAA: OR 1.54 95% CI 1.03– 2.32;

214

E.-C. Tan et al. / Molecular Brain Research 139 (2005) 212 – 216

Table 1 Distribution of genotype and allele frequency of the two insertion/deletion polymorphisms in the nogo gene Group

Genotype

Allele

CAA/CAA All cases (363) All control (253)

CAA/

/

43 (0.12) 165 (0.45) 25 (0.10) 100 (0.40) v 2 = 3.761, P = 0.153 ( P c = 0.306) 12 (0.13) 43 (0.46) 7 (0.05) 55 (0.43) v 2 = 5.091, P = 0.078 ( P c = 0.156) 31 (0.12) 122 (0.45) 18 (0.15) 45 (0.36) v 2 = 2.861, P = 0.239 ( P c = 0.478)

Female cases (93) Female controls (129) Male cases (270) Male controls (124)

TATC/TATC All cases (363) All control (253)

38 (0.41) 67 (0.52) 117 (0.43) 61 (0.49)

TATC/

/

137 (0.38) 171 (0.47) 116 (0.46) 104 (0.41) v 2 = 4.053, P = 0.132 ( P c = 0.164) 32 (0.34) 46 (0.49) 63 (0.49) 55 (0.43) v 2 = 5.849, P = 0.054 ( P c = 0.108) 105 (0.39) 125 (0.46) 53 (0.43) 49 (0.40) v 2 = 1.662, P = 0.436 ( P c = 0.872)

Female cases Female controls Male cases Male controls

TATC: OR 1.62 95% CI 1.09 –2.41). Comparison between female patients and controls also revealed marginally significant association at the genotypic level for both polymorphisms (Table 1). When P values were corrected for multiple comparisons, the allelic difference between cases and controls for females remained significant when the alpha value was set at 5% ( P c in Table 1). There was almost complete linkage disequilibrium between the two insertion/deletion polymorphisms with the CAA insertion and TATC deletion forming one common haplotype and the wildtype CAA and TATCTATC being the other (Table 2). Out of the 827 chromosomes carrying the wildtype CAA, only 6% (21 from controls and 30 from cases) did not have the wildtype TATCTATC as well. And of the 401 chromosomes with the CAA insertion, all but 4 (1%) also carry the TATC deletion. Linkage disequilibrium as calculated by the EM algorithm gave the value of 0.988 for D’ and 0.906 for r. Analysis with haplotypes did not reveal statistically significant association of any haplotype with schizophrenia

CAA

155 (0.43) 128 (0.50)

0.346 0.654 0.296 0.704 v 2 = 3.299, P = 0.069 ( P c = 0.138) 0.360 0.640 0.267 0.733 v 2 = 4.378, P = 0.036 ( P c = 0.072) 0.341 0.659 0.327 0.673 v 2 = 0.152, P = 0.697 ( P c = 0.304) TATC

55 (0.15) 33 (0.13) 15 (0.16) 11 (0.08) 40 (0.15) 22 (0.17)

0.613 0.387 0.664 0.336 v 2 = 3.353, P = 0.067 ( P c = 0.134) 0.591 0.409 0.702 0.298 v 2 = 5.807, P = 0.016 ( P c = 0.032) 0.620 0.380 0.625 0.375 v 2 = 0.016, P = 0.901 ( P c = 1.000)

(v 2 = 0.015 – 3.354, df = 3, P = 0.067 –0.902) for all patients combined. When subjects were separated according to gender, there was significant association for the female subgroup. The two haplotypes containing the CAA insertion were at increased risk for schizophrenia while the haplotypes with the wildtype CAA had a protective effect (Table 2). As our control subjects are significantly younger than the cases, we also reanalyzed the data after removing those below 30 years of age, as they might still develop schizophrenia after the point of recruitment into the study. With only 135 subjects left in the control group, there was no statistically significant association in either genotype or allele frequency for both markers between cases and controls for all analysis groups.

4. Discussion It has been reported that there was higher prevalence of the CAA insertion in patients with schizophrenia [15]. Our

Table 2 Distribution of NOGO 3VUTR CAA/TATC haplotypes in controls and patients with schizophrenia Wildtypea

All

CAA

TATC

+ +

+ +

Females only

+ +

+ +

a

Wildtype is 1 copy of CAA and 2 copies of TATC.

Controls (n)

Cases (n)

v2

P value ( P c)

OR

21 334 150 3 9 177 69 3

30 442 251 1 9 110 67 0

0.015 2.878 3.354

0.902 (1.804) 0.090 (0.180) 0.067 (0.134)

0.99 (0.56, 1.76) 0.81 (0.64, 1.03) 1.27 (0.99, 1.62)

0.211 4.915 4.340

0.646 (1.292) 0.027 (0.054) 0.037 (0.074)

1.40 (0.55, 3.60) 0.63 (0.42, 0.93) 1.57 (1.05, 2.37)

E.-C. Tan et al. / Molecular Brain Research 139 (2005) 212 – 216

results are consistent with an excess of the CAA inserts at both the allelic and genotypic level. However, statistically significant difference was only achieved for the female subgroup. A study from China on these two insertion/ deletion polymorphisms and another two single nucleotide polymorphisms in the promoter of the gene did not find association with all 4 markers [3]. Another study on populations of French-Canadian and Tunisian origins found higher frequency of homozygous CAA insertion in cases but the difference did not reach statistical significance [21]. The positive association was not replicated by Covault et al., who observed population difference in the frequency of the CAA insertion at 0.43 for European Americans and 0.20 for African Americans [5]. Our allele frequency for the CAA insertion is in-between at 0.30 for control subjects of Chinese ancestry. However, it was reported to be 0.685 for cases and 0.664 for the controls by Chen et al. [3]. That would make it the major allele and the reverse of all other studies carried out thus far, including our population which is also of Chinese descent. We next studied the joint effects of the two insertion/ deletion polymorphisms. Within the female group, although there was positive association of each of the two markers with schizophrenia, combining the two markers did not significantly increase the predictive risk for schizophrenia. The odds ratio for the haplotype containing the two at-risk alleles is the average of the two individual markers. It also appears that the TATC polymorphism has the stronger association with schizophrenia. Although the results of our study do not directly confirm the involvement of the homozygous CAA insertion genotype in schizophrenia, we found statistically significant association of the TATC deletion and CAA insertion allelic frequencies with chronic schizophrenia in females. Given the evidence that nogo mRNA level is altered in schizophrenia, its involvement in the pathogenesis cannot be ruled out. Alterations in myelination, myelin-related dysfunction, and levels of oligodendrocyte and myelin proteins such as myelin-associated glycoprotein (MAG) and myelin basic protein (MBP) have all been documented in schizophrenia [6,7,8,20]. As Nogo is a myelin-derived axonal regeneration inhibitor, one mechanism by which the insertion/deletion polymorphisms may result in schizophrenia is that one or both, or another polymorphism that is in linkage disequilibrium with them might cause abnormal expression of the protein. Aberrant expression could disturb the normal repair process and cause aberrant neuronal connections, which might result in the symptoms of schizophrenia. Indeed, application of antibodies against Nogo-A would induce cell body changes and axonal sprouting in intact Purkinje cells [18], and Nogo-C has also been shown to block axonal regeneration in transgenic mice [12]. Aberrant neuronal connectivity and information processing have also been reported in mice that are deficient in the CHL1 (close homolog of L1) gene, a gene implicated in mental impair-

215

ment and symptoms of schizophrenia in the 3p-syndrome [9,14]. Difference in the patterns of functional neuronal connectivity between controls and subjects with schizophrenia has been demonstrated using positron emission tomography (PET) followed by neural network analysis [11]. Taking the molecular finding together with the present genetic findings, the nogo gene appears to be a credible candidate gene related to the susceptibility to schizophrenia for females. Further genetic studies with other populations are needed to support our data. Molecular studies are also needed to unravel the functional significance of the two polymorphisms and to uncover any true causative variant if neither of the insertion/deletion polymorphisms turns out to be the causative one.

Acknowledgments This research was supported in part by the National Medical Research Council Block Grant S035 to the Institute of Mental Health and the Human Genetics Program of the Defence Science and Technology Agency.

References [1] N.J. Camp, S.L. Neuhausen, J. Tiobech, A. Polloi, H. Coon, M. Myles-Worsley, Genomewide multipoint linkage analysis of seven extended Palauan pedigrees with schizophrenia, by a Markovchain Monte Carlo method, Am. J. Hum. Genet. 69 (2001) 1278 – 1289. [2] M.S. Chen, A.B. Huber, M.E. van der Haar, M. Frank, L. Schnell, A.A. Spillmann, F. Christ, M.E. Schwab, Nogo-A is a myelinassociated neurite outgrowth inhibitor and an antigen for monoclonal antibody IN-1, Nature 403 (2000) 434 – 439. [3] W. Chen, N. Gu, S. Duan, Y. Sun, Y. Zheng, C. Li, Y. Pan, Y. Xu, G. Feng, L. He, No association between the genetic polymorphisms within RTN4 and schizophrenia in the Chinese population, Neurosci. Lett. 365 (2004) 23 – 27. [4] H. Coon, M. Myles-Worsley, J. Tiobech, M. Hoff, J. Rosenthal, P. Bennett, F. Reimherr, P. Wender, P. Dale, A. Polloi, W. Byerley, Evidence for a chromosome 2p13 – 14 schizophrenia susceptibility locus in families from Palau, Micronesia, Mol. Psychiatry 3 (1998) 521 – 527. [5] J. Covault, J. Lee, K. Jensen, H. Kranzler, Nogo 3V-untranslated region CAA insertion: failure to replicate association with schizophrenia and demonstration of marked population difference in frequency of the insertion, Brain Res. Mol. Brain Res. 120 (2004) 197 – 200. [6] K.L. Davis, D.G. Stewart, J.I. Friedman, M. Buchsbaum, P.D. Harvey, P.R. Hof, J. Buxbaum, V. Haroutunian, White matter changes in schizophrenia: evidence for myelin-related dysfunction, Arch. Gen. Psychiatry 60 (2003) 443 – 456. [7] S.W. Flynn, D.J. Lang, A.L. Mackay, V. Goghari, I.M. Vavasour, K.P. Whittall, G.N. Smith, V. Arango, J.J. Mann, A.J. Dwork, P. Falkai, W.G. Honer, Abnormalities of myelination in schizophrenia detected in vivo with MRI, and post-mortem with analysis of oligodendrocyte proteins, Mol. Psychiatry 8 (2003) 811 – 820. [8] P.R. Hof, V. Haroutunian, C. Copland, K.L. Davis, J.D. Buxbaum, Molecular and cellular evidence for an oligodendrocyte abnormality in schizophrenia, Neurochem. Res. 27 (2002) 1193 – 1200. [9] A. Irintchev, M. Koch, L.K. Needham, P. Maness, M. Schachner, Impairment of sensorimotor gating in mice deficient in the cell

216

[10]

[11]

[12] [13]

[14]

[15]

[16]

E.-C. Tan et al. / Molecular Brain Research 139 (2005) 212 – 216 adhesion molecule L1 or its close homologue, CHL1, Brain Res. 1029 (2004) 131 – 134. A. Josephson, J. Widenfalk, H.W. Widmer, L. Olson, C. Spenger, NOGO mRNA expression in adult and fetal human and rat nervous tissue and in weight drop injury, Exp. Neurol. 169 (2001) 319 – 328. G.M. Jossin, P.F. Liddle, Neural network analysis of the pattern of functional connectivity between cerebral areas in schizophrenia, Biol. Cybernetics 84 (2001) 117 – 122. J.E. Kim, I.E. Bonilla, D. Qiu, S.M. Strittmatter, Nogo-C is sufficient to delay nerve regeneration, Mol. Cell. Neurosci. 23 (2003) 451 – 459. C.M. Lewis, D.F. Levinson, L.H. Wise, L.E. DeLisi, R.E. Straub, I. Hovatta, N.M. Williams, S.G. Schwab, A.E. Pulver, S.V. Faraone, et al., Genome scan meta-analysis of schizophrenia and bipolar disorder: Part II. Schizophrenia, Am J. Hum. Genet. 73 (2003) 34 – 48. M. Montag-Sallaz, A. Baarke, D. Montag, Aberrant neuronal connectivity in CHL1-deficient mice is associated with altered information processing-related immediate early gene expression, J. Neurobiol. 57 (2003) 67 – 80. G. Novak, D. Kim, P. Seeman, T. Tallerico, Schizophrenia and Nogo: elevated mRNA in cortex, and high prevalence of a homozygous CAA insert, Brain Res. Mol. Brain Res. 107 (2002) 183 – 189. T. Oertle, C. Huber, M. van der Putten, M. Schwab, Genomic structure and functional characterisation of the promoters of human and mouse nogo/rtn4, J. Mol. Biol. 325 (2003) 299 – 323.

[17] M. Reindl, S. Khantane, R. Ehling, K. Schanda, A. Lutterotti, C. Brinkhoff, T. Oertle, M.E. Schwab, F. Deisenhammer, T. Berger, C.E. Bandtlow, Serum and cerebrospinal fluid antibodies to Nogo-A in patients with multiple sclerosis and acute neurological disorders, J. Neuroimmunol. 145 (2003) 139 – 147. [18] F. Rossi, A. Buffo, P. Strata, Regulation of intrinsic regenerative properties and axonal plasticity in cerebellar Purkinje cells, Restor. Neurol. Neurosci. 19 (2001) 85 – 94. [19] S.H. Shaw, M. Kelly, A.B. Smith, G. Shields, P.J. Hopkins, J. Loftus, S.H. Laval, A. Vita, M. De Hert, L.R. Cardon, T.J. Crow, R. Sherrington, L.E. DeLisi, A genome-wide search for schizophrenia susceptibility genes, Am. J. Med. Genet. 81 (1998) 364 – 376. [20] D. Tkachev, M.L. Mimmack, M.M. Ryan, M. Wayland, T. Freeman, P.B. Jones, M. Starkey, M.J. Webster, R.H. Yolken, S. Bahn, Oligodendrocyte dysfunction in schizophrenia and bipolar disorder, Lancet 362 (2003) 798 – 805. [21] L. Xiong, G.A. Rouleau, L.E. Delisi, J. St-Onge, R. Najafee, J.B. Riviere, C. Benkelfat, K. Tabbane, F. Fathalli, Z. Danics, A. Labelle, S. Lal, R. Joober, CAA insertion polymorphism in the 3VUTR of Nogo gene on 2p14 is not associated with schizophrenia, Brain Res. Mol. Brain Res. 133 (2005) 153 – 156. [22] J. Yang, L. Yu, A.D. Bi, S.Y. Zhao, Assignment of the human reticulon 4 gene (RTN4) to chromosome 2p14Y2p13 by radiation hybrid mapping, Cytogenet. Cell Genet. 88 (2000) 101 – 102.