Journal of Neuroimmunology 279 (2015) 46–49
Contents lists available at ScienceDirect
Journal of Neuroimmunology journal homepage: www.elsevier.com/locate/jneuroim
Short communication
MS susceptibility is not affected by single nucleotide polymorphisms in the MMP9 gene S. Nischwitz a,⁎,1, C. Wolf a,1, T.F.M. Andlauer a, D. Czamara a, U.K. Zettl b, P. Rieckmann c, D. Buck d, M. Ising a, T. Bettecken a, B. Mueller-Myhsok a, F. Weber a a
Max Planck Institute of Psychiatry, Kraepelinstr. 2-10, 80804 Munich, Germany Department of Neurology, University of Rostock, Gehlsheimer Straße 20, 18147 Rostock, Germany Department of Neurology, Sozialstiftung Bamberg, Buger Straße 80, 96049 Bamberg, Germany d Department of Neurology, TU München, Ismaninger Str. 22, 81675 München, Germany b c
a r t i c l e
i n f o
Article history: Received 1 December 2014 Received in revised form 21 January 2015 Accepted 22 January 2015 Keywords: Multiple sclerosis MMP9 Susceptibility GWAS Imputation
a b s t r a c t Matrix metalloproteinase 9 (MMP9) plays an important role in the pathogenesis of multiple sclerosis (MS). However, the impact of genetic variants affecting MMP9 on MS susceptibility is still in debate. We could not detect an association of MMP9 SNPs with MS on a genome-wide significance level by SNP genotyping, followed by imputation of SNPs within a region stretching 2 Mbp up- and down-stream of MMP9. Rs6073751, located within WFDC2, was found associated with MS most strongly. Rs3918242, associated with MS according to previous reports, showed nominal significance only. Meta-analysis of our own and published data did not confirm this effect. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Several studies have demonstrated the importance of MMP9 in the pathophysiology of multiple sclerosis (MS). At least in the early stages, MS is characterized by an invasion of T- and B-lymphocytes as well as of monocytes into the tissues of the central nervous system. MMP9 is expressed in these disease-relevant immune cells. It is involved in the following processes associated with MS: (1) breakdown of the blood– brain barrier, (2) opening of the basal lamina during inflammation, and (3) cleavage of myelin basic protein and thereby production of potentially immunogenic oligopeptides (Chandler et al., 1997; Kouwenhoven et al., 2001, 2002; Abraham et al., 2005; Dressel et al., 2007; Shiryaev et al., 2009). An elevation of MMP9 levels in serum and cerebrospinal fluid (CSF) of MS patients correlates positively with disease activity and inversely with treatment (Sellebjerg et al., 2000; Bernal et al., 2009). However, the impact of genetic variants of the MMP9 gene on MS susceptibility is still controversial. Previously, candidate gene approaches focusing on single nucleotide polymorphisms (SNPs) were conducted using limited sample sizes and produced inconsistent results (Table 1). In general, genome-wide association studies (GWAS) are im⁎ Corresponding author. E-mail address: [email protected] (S. Nischwitz). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.jneuroim.2015.01.008 0165-5728/© 2015 Elsevier B.V. All rights reserved.
peded because the genotyping arrays typically used cover the target region with a low SNP density only. Our aims were (a) to improve the statistical power by genotyping a large sample size, (b) to nominally increase the SNP coverage of the MMP9 region and the flanking +/−2 mega base pairs (Mbp), a region containing more than 70 genes, by imputation of additional variants, and (c) to perform a meta-analysis on our own findings together with published data.
2. Materials and methods We genotyped 3002 individuals (1652 cases and 1350 controls) of German origin using Illumina 300 K, 370 K and 660 K genotyping arrays. All patients fulfilled the McDonald criteria for MS diagnosis. Informed written consent was obtained from all individuals. The ethics committee of the Medical Faculty at the Ludwig Maximilian University, Munich and the local ethic committees of the participating study centres have approved the study. Complete data, including age and gender, were available for 2565 individuals (Table 2). We determined the first five axes of variation resulting from a multidimensional scaling analysis (MDS) of the identity by state (IBS) matrix. These were used as covariates in the analysis, thereby taking population stratification into account. SNPs with a minor allele frequency (MAF) b0.01, a call rate b98%, or a significant deviation from Hardy–Weinberg-Equilibrium
S. Nischwitz et al. / Journal of Neuroimmunology 279 (2015) 46–49
47
Table 1 Published findings concerning rs3918242, formerly known as −1562C/T. The published results may differ from the data used in the meta-analysis, which was restricted to the dominant genetic model. Reference
p-Value (corrected for multiple testing)
Odds ratio (OR) or χ2
Effect on MS susceptibility
Cases and controls Frequencies T-carrier
Population
Nelissen et al., 2000
Not available
Not available
No effect on MS susceptibility
Not available
Not available
Benesová et al., 2008
p = 0.01 pcorr = 0.05
OR 0.58 95% CI: 0.38–0.89
244 cases, 132 controls
MS: 53/244 (21.7%) Controls: 45/132 (34.1%)
Czech
Mirowska-Guzel et al., 2009 Fernandes et al., 2009
p = 0.0030
La Russa et al., 2010
p = 5.6 × 10−5 pcorr = 4.0 × 10−4
OR 2.57 95% CI: 1.52–4.36
234 cases, 190 controls 158 cases, 191 controls 243 cases, 173 controls
MS: 106/234 (45.3%) Controls: 54/190 (28.4%) MS: 41/158 (25.9%) Controls: 35/191 (18.3%) MS: 79/243 (32.5%) Controls: 26/173 (15.0%)
Polish
p = 0.12
OR 1.7 95% CI: 1.2–2.4 x2 = 2.38
No significant decrease of T alleles in all MS patients, but significant decrease in female patients with MS compared to healthy female controls (χ2 = 6.12, df = 1, p = 0.01) Nominal significant decrease of T alleles in MS patients. T allele less frequent in female MS patients Significant increase of T alleles in MS patients
MS: 56/199 (28,1%) Controls: 44/146 (30.1%) MS: 41/187 (21.9%) Controls: 82/282 (29.1%)
Swedish
Zivković et al., 2007
199 cases, 146 controls 187 cases, 282 controls
No significant increase of T alleles in MS patients Significant increase of T alleles in MS patients
(HWE) with p b 10− 6 were excluded. All samples were merged and only SNPs genotyped in all samples were used for further processing. For the imputation, we first prephased the genotype data using SHAPEIT (Delaneau et al., 2011), and subsequently imputed SNPs within the MMP9 gene and the two flanking 2 Mbp regions (hg19:chr20: 42,637,658-46,645,154) to the 1000 genomes references (phase I integrated variant set release (v3) in NCBI build 37 (hg19)) using IMPUTE2 v2.3.0 (Marchini et al., 2007). SNPs with an imputation quality b0.8 (IMPUTE2 info score) and a MAF b1% were excluded. For the final analysis, genotypes of 1421 cases and 1345 controls for 55,884 SNPs were used. On these data, we performed logistic regression using the software tool PLINK v1.07 (Purcell et al., 2007), including age, sex and the first five axes of variation from the MDS analysis as covariates (dosage, allelic model). Only results with an R2 quality metric/information content N 0.9 and b 1.0 were included in downstream analyses (8303 SNPs). Finally, we performed a random-effects meta-analysis on our own imputed and previously published data using R 2.15.2 (http://www.rproject.org/) and the package rmeta. Based on published genotype distributions, ORs and CIs were calculated for the dominant model without covariates. The power of each study was determined using the Genetic Power Calculator (Purcell et al., 2003). 3. Results We performed dosage analyses using an allelic model and including age, sex, and the first five axes of variation from the MDS analysis as covariates. None of these SNPs showed a significant association with MS after correction for multiple testing (Bonferroni adjusted significance level α = 8.95 × 10−7, Fig. 1A). SNP rs6073751 demonstrated the strongest effect with a nominal p-value of 0.00121. This SNP, however, is not located within the MMP9 gene, but in the upstream gene WAP fourdisulfide core domain 2 (WFDC2). At MMP9, the SNP rs3918242 (Fig. 1A,B), which has been previously reported as associated with MS (Table 1), showed a nominally significant effect with a p-value of
Table 2 Demographic data of cases and controls.
Total number Females Males Mean age [years] (+/−standard deviation) Age range
Cases
Controls
1421 70.4% 29.6% 39.3 (+/−10.22) 17–81
1345 62.6% 37.4% 50.0 (+/−13.46) 18–90
Serbian
Brazilian Southern Italian
0.02936 (OR referring to the minor allele (T) = 0.82, CI = 0.69–0.98). This suggests a decreased frequency of the T allele in our MS patients. Separate calculations for females and males revealed a greater reduction of T-alleles in female MS patients in our cohort (OR 1.26 vs 1.16). A meta-analysis using a random-effects model, including six previously published studies and our own imputed data, showed no significant association between SNPs within the MMP9 gene and MS susceptibility when applying a dominant genetic model (Fig. 2). We used a random-effects model, as there was evidence for heterogeneity between the studies (p-value for heterogeneity = 4.7 × 10−7). 4. Discussion The aim of our study was to elucidate the genetic impact of variants within the MMP-9 gene and its two flanking 2 Mbp regions on MS susceptibility. Rs6073751, located within WFDC2, demonstrated the strongest effect. Although it did not remain significant after correction for multiple testing, WFDC2 may be important in the pathophysiology of MS, as it is part of the WAP domain protein family known to modulate immunity (Makarycheva et al., 2011, Wilkinson et al., 2011). Recently, interpopulation heterogeneity of the WFDC gene region has been demonstrated, probably driven by the diversity of pathogens on different continents (Ferreira et al., 2013). This finding is interesting, as MS prevalence decreases from the poles to the equator. Thus, further studies on the regional variants of these genes and their role on autoimmunity are warranted. Subsequently, we focused on rs3918242 (formerly known as − 1562C/T), located close to MMP9, and previously reported as being associated with MS. We found a nominally significant effect towards a decreased T allele frequency in German MS patients, similar to the results by Benesová et al. (2008) observed in a Czech population. However, this finding did not remain significant after correction for multiple testing and is discrepant to results from Southern Italian, Polish, and Brazilian samples, which exhibited an increased T allele frequency in MS patients (Fernandes et al., 2009; Mirowska-Guzel et al., 2009; La Russa et al., 2010). Previous reports have suggested that the T allele is underrepresented only in female MS patients (Zivković et al., 2007 and Benesová et al., 2008). Although the effect on MS susceptibility is not significant, neither in males nor in females, the direction of our results is in line with the data of Benesová et al. and Zivković et al. We performed a meta-analysis on our data together with the data of the studies that had investigated this SNP previously. We found no overall significant effect of this SNP on MS susceptibility. Moreover, based on our German sample, these results do not support common genetic variants in the MMP9 gene and its flanking +/− 2 Mbp region to be risk factors for MS. This is in line with the results of a recently published Australian study and a linkage
48
S. Nischwitz et al. / Journal of Neuroimmunology 279 (2015) 46–49
Fig. 1. A: No SNP within MMP9 or +/−2 Mbp of the gene showed a significant association with MS after correction for multiple testing (Bonferroni adjusted significance level α = 8.95 × 10−7). SNP rs6073751 is labelled in green, rs3918242 in red. B: At the locus MMP9, the SNP rs3918242, which has previously been found associated with MS susceptibility, showed a nominally significant p-value of 0.02936, indicating a decrease of the T-allele in our cohort of German MS patients. However, there was no significant association after correction for multiple testing.
analysis in 100 Russian families (Makarycheva et al., 2011; Cox et al., 2014). Both studies, however, could not be included in the metaanalysis for methodological reasons.
In conclusion, our results do not suggest that MS susceptibility is affected by known genetic variants within MMP9, although MMP9 itself may play an important role in MS pathophysiology.
Fig. 2. Meta-analysis of previously published data and our results on rs3918242 (formerly known as −1562C/T) located upstream to the MMP9 gene, using a dominant model. “Power” indicates the power to detect an association in a dominant model with a nominal p-value of 0.05, using an average minor allele frequency of 0.148 and the ORs indicated for each study (Purcell, 2003).
S. Nischwitz et al. / Journal of Neuroimmunology 279 (2015) 46–49
Statement of conflicts of interests S. Nischwitz received travel grants for the attention of scientific meetings from Merck Serono and TEVA and grant support from Bayer-Schering. P Rieckmann received honoraria for lectures and steering committee meetings from Baxter, Bayer, Biogen Idec, Boehringer-Ingelheim, Bristol-Myers Squibb, Genpharm, Genzyme, Merck KG, Novartis, Pfizer, Roche, Sanofi-Aventis, Siemens, and Teva pharmaceutical Industries. UK Zettl received travel grants for the attention of scientific meetings and honoraria for speaking from Biogen-Idec, Merck-Serono, Bayer, TEVA, Sanofi-Genzyme, Almirall and Novartis. D Buck received compensation for activities with Bayer HealthCare, Biogen Idec, Merck Serono, and Novartis. She was supported by the Commission for Clinical Research of the Faculty of Medicine, Technical University Munich, Abirisk, and the PML Consortium. F Weber received honoraria from Genzyme and Novartis for serving on a scientific advisory board and a travel grant for the attention of a scientific meeting from Merck-Serono, Biogen Idec and Novartis and received grant support from Merck-Serono and the Federal Ministry of Education and Research (BMBF, Projects Biobanking and Omics in ControlMS as part of the Competence Network Multiple Sclerosis). C Wolf, TFM Andlauer, D Czamara, M Ising, T Bettecken and B Mueller-Myhsok report no disclosures.
References Abraham, M., Shapiro, S., Karni, A., Weiner, H.L., Miller, A., 2005. Gelatinases (MMP-2 and MMP-9) are preferentially expressed by Th1 vs. Th2 cells. J. Neuroimmunol. 163, 157–164. Benesová, Y., Vasků, A., Stourac, P., Hladíková, M., Beránek, M., Kadanka, Z., Novotná, H., Bednarík, J., 2008. Matrix metalloproteinase-9 and matrix metalloproteinase-2 gene polymorphisms in multiple sclerosis. J. Neuroimmunol. 205, 105–109. Bernal, F., Elias, B., Hartung, H.P., Kieseier, B.C., 2009. Regulation of matrix metalloproteinases and their inhibitors by interferon-beta: a longitudinal study in multiple sclerosis patients. Mult. Scler. 15, 721–727. Chandler, S., Miller, K.M., Clements, J.M., Lury, J., Corkill, D., Anthony, D.C., Adams, S.E., Gearing, A.J., 1997. Matrix metalloproteinases, tumor necrosis factor and multiple sclerosis: an overview. J. Neuroimmunol. 72, 155–161. Cox, M.B., Bowden, N.A., Scott, R.J., Lechner-Scott, J., 2014. Common genetic variants in the plasminogen activation pathway are not associated with multiple sclerosis. Mult. Scler. 20, 489–491. Delaneau, O., Marchini, J., Zagury, J.F., 2011. A linear complexity phasing method for thousands of genomes. Nat. Methods 9, 179–181.
49
Dressel, A., Mirowska-Guzel, D., Gerlach, C., Weber, F., 2007. Migration of T-cell subsets in multiple sclerosis and the effect of interferon-beta1a. Acta Neurol. Scand. 116, 164–168. Fernandes, K.S., Brum, D.G., Sandrim, V.C., Guerreiro, C.T., Barreira, A.A., Tanus-Santos, J.E., 2009. Matrix metalloproteinase-9 genotypes and haplotypes are associated with multiple sclerosis and with the degree of disability of the disease. J. Neuroimmunol. 214, 128–131. Ferreira, Z., Seixas, S., Andrés, A.M., Kretzschmar, W.W., Mullikin, J.C., Cherukuri, P.F., Cruz, P., Swanson, W.J., Comparative Sequencing Program, N.I.S.C., Clark, A.G., Green, E.D., Hurle, B., 2013. Reproduction and immunity-driven natural selection in the human WFDC locus. Mol. Biol. Evol. 30, 938–950. Kouwenhoven, M., Ozenci, V., Gomes, A., Yarilin, D., Giedraitis, V., Press, R., Link, H., 2001. Multiple sclerosis: elevated expression of matrix metalloproteinases in blood monocytes. J. Autoimmun. 16, 463–470. Kouwenhoven, M., Ozenci, V., Tjernlund, A., Pashenkov, M., Homman, M., Press, R., Link, H., 2002. Monocyte-derived dendritic cells express and secrete matrix-degrading metalloproteinases and their inhibitors and are imbalanced in multiple sclerosis. J. Neuroimmunol. 126, 161–171. La Russa, A., Cittadella, R., De Marco, E.V., Valentino, P., Andreoli, V., Trecroci, F., Latorre, V., Gambardella, A., Quattrone, A., 2010. Single nucleotide polymorphism in the MMP-9 gene is associated with susceptibility to develop multiple sclerosis in an Italian case– control study. J. Neuroimmunol. 225, 175–179. Makarycheva, O.Y., Tsareva, E.Y., Sudomoina, M.A., Kulakova, O.G., Titov, B.V., Bykova, O.V., Gol'tsova, N.V., Kuzenkova, L.M., Boiko, A.N., Favorova, O.O., 2011. Family analysis of linkage and association of HLA-DRB1, CTLA4, TGFB1, IL4, CCR5, RANTES, MMP9 and TIMP1 gene polymorphisms with multiple sclerosis. Acta Nat. 3, 85–92. Marchini, J., Howie, B., Myers, S., McVean, G., Donnelly, P., 2007. A new multipoint method for genome-wide association studies via imputation of genotypes. Nat. Genet. 39, 906–913. Mirowska-Guzel, D., Gromadzka, G., Czlonkowski, A., Czlonkowska, A., 2009. Association of MMP1, MMP3, MMP9, and MMP12 polymorphisms with risk and clinical course of multiple sclerosis in a Polish population. J. Neuroimmunol. 214, 113–117. Nelissen, I., Vandenbroeck, K., Fiten, P., Hillert, J., Olsson, T., Marrosu, M.G., Opdenakker, G., 2000. Polymorphism analysis suggests that the gelatinase B gene is not a susceptibility factor for multiple sclerosis. J. Neuroimmunol. 105, 58–63. Purcell, S., Cherny, S.S., Sham, P.C., 2003. Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics 19, 149–150. Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A., Bender, D., Maller, J., Sklar, P., de Bakker, P.I., Daly, M.J., Sham, P.C., 2007. PLINK: a toolset for whole-genome association and population-based linkage analysis. Am. J. Hum. Genet. 81, 559–575. Sellebjerg, F., Madsen, H.O., Jensen, C.V., Jensen, J., Garred, P., 2000. CCR5 delta32, matrix metalloproteinase-9 and disease activity in multiple sclerosis. J. Neuroimmunol. 102, 98–106. Shiryaev, S.A., Savinov, A.Y., Cieplak, P., Ratnikov, B.I., Motamedchaboki, K., Smith, J.W., Strongin, A.Y., 2009. Matrix metalloproteinase proteolysis of the myelin basic protein isoforms is a source of immunogenic peptides in autoimmune multiple sclerosis. PLoS One 4, e4952. Wilkinson, T.S., Roghanian, A., Simpson, A.J., Sallenave, J.M., 2011. WAP domain proteins as modulators of mucosal immunity. Biochem. Soc. Trans. 39, 1409–1415. Zivković, M., Djurić, T., Dincić, E., Raicević, R., Alavantić, D., Stanković, A., 2007. Matrix metalloproteinase-9 −1562 C/T gene polymorphism in Serbian patients with multiple sclerosis. J. Neuroimmunol. 189, 147–150.
Contents lists available at ScienceDirect
Journal of Neuroimmunology journal homepage: www.elsevier.com/locate/jneuroim
Short communication
MS susceptibility is not affected by single nucleotide polymorphisms in the MMP9 gene S. Nischwitz a,⁎,1, C. Wolf a,1, T.F.M. Andlauer a, D. Czamara a, U.K. Zettl b, P. Rieckmann c, D. Buck d, M. Ising a, T. Bettecken a, B. Mueller-Myhsok a, F. Weber a a
Max Planck Institute of Psychiatry, Kraepelinstr. 2-10, 80804 Munich, Germany Department of Neurology, University of Rostock, Gehlsheimer Straße 20, 18147 Rostock, Germany Department of Neurology, Sozialstiftung Bamberg, Buger Straße 80, 96049 Bamberg, Germany d Department of Neurology, TU München, Ismaninger Str. 22, 81675 München, Germany b c
a r t i c l e
i n f o
Article history: Received 1 December 2014 Received in revised form 21 January 2015 Accepted 22 January 2015 Keywords: Multiple sclerosis MMP9 Susceptibility GWAS Imputation
a b s t r a c t Matrix metalloproteinase 9 (MMP9) plays an important role in the pathogenesis of multiple sclerosis (MS). However, the impact of genetic variants affecting MMP9 on MS susceptibility is still in debate. We could not detect an association of MMP9 SNPs with MS on a genome-wide significance level by SNP genotyping, followed by imputation of SNPs within a region stretching 2 Mbp up- and down-stream of MMP9. Rs6073751, located within WFDC2, was found associated with MS most strongly. Rs3918242, associated with MS according to previous reports, showed nominal significance only. Meta-analysis of our own and published data did not confirm this effect. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Several studies have demonstrated the importance of MMP9 in the pathophysiology of multiple sclerosis (MS). At least in the early stages, MS is characterized by an invasion of T- and B-lymphocytes as well as of monocytes into the tissues of the central nervous system. MMP9 is expressed in these disease-relevant immune cells. It is involved in the following processes associated with MS: (1) breakdown of the blood– brain barrier, (2) opening of the basal lamina during inflammation, and (3) cleavage of myelin basic protein and thereby production of potentially immunogenic oligopeptides (Chandler et al., 1997; Kouwenhoven et al., 2001, 2002; Abraham et al., 2005; Dressel et al., 2007; Shiryaev et al., 2009). An elevation of MMP9 levels in serum and cerebrospinal fluid (CSF) of MS patients correlates positively with disease activity and inversely with treatment (Sellebjerg et al., 2000; Bernal et al., 2009). However, the impact of genetic variants of the MMP9 gene on MS susceptibility is still controversial. Previously, candidate gene approaches focusing on single nucleotide polymorphisms (SNPs) were conducted using limited sample sizes and produced inconsistent results (Table 1). In general, genome-wide association studies (GWAS) are im⁎ Corresponding author. E-mail address: [email protected] (S. Nischwitz). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.jneuroim.2015.01.008 0165-5728/© 2015 Elsevier B.V. All rights reserved.
peded because the genotyping arrays typically used cover the target region with a low SNP density only. Our aims were (a) to improve the statistical power by genotyping a large sample size, (b) to nominally increase the SNP coverage of the MMP9 region and the flanking +/−2 mega base pairs (Mbp), a region containing more than 70 genes, by imputation of additional variants, and (c) to perform a meta-analysis on our own findings together with published data.
2. Materials and methods We genotyped 3002 individuals (1652 cases and 1350 controls) of German origin using Illumina 300 K, 370 K and 660 K genotyping arrays. All patients fulfilled the McDonald criteria for MS diagnosis. Informed written consent was obtained from all individuals. The ethics committee of the Medical Faculty at the Ludwig Maximilian University, Munich and the local ethic committees of the participating study centres have approved the study. Complete data, including age and gender, were available for 2565 individuals (Table 2). We determined the first five axes of variation resulting from a multidimensional scaling analysis (MDS) of the identity by state (IBS) matrix. These were used as covariates in the analysis, thereby taking population stratification into account. SNPs with a minor allele frequency (MAF) b0.01, a call rate b98%, or a significant deviation from Hardy–Weinberg-Equilibrium
S. Nischwitz et al. / Journal of Neuroimmunology 279 (2015) 46–49
47
Table 1 Published findings concerning rs3918242, formerly known as −1562C/T. The published results may differ from the data used in the meta-analysis, which was restricted to the dominant genetic model. Reference
p-Value (corrected for multiple testing)
Odds ratio (OR) or χ2
Effect on MS susceptibility
Cases and controls Frequencies T-carrier
Population
Nelissen et al., 2000
Not available
Not available
No effect on MS susceptibility
Not available
Not available
Benesová et al., 2008
p = 0.01 pcorr = 0.05
OR 0.58 95% CI: 0.38–0.89
244 cases, 132 controls
MS: 53/244 (21.7%) Controls: 45/132 (34.1%)
Czech
Mirowska-Guzel et al., 2009 Fernandes et al., 2009
p = 0.0030
La Russa et al., 2010
p = 5.6 × 10−5 pcorr = 4.0 × 10−4
OR 2.57 95% CI: 1.52–4.36
234 cases, 190 controls 158 cases, 191 controls 243 cases, 173 controls
MS: 106/234 (45.3%) Controls: 54/190 (28.4%) MS: 41/158 (25.9%) Controls: 35/191 (18.3%) MS: 79/243 (32.5%) Controls: 26/173 (15.0%)
Polish
p = 0.12
OR 1.7 95% CI: 1.2–2.4 x2 = 2.38
No significant decrease of T alleles in all MS patients, but significant decrease in female patients with MS compared to healthy female controls (χ2 = 6.12, df = 1, p = 0.01) Nominal significant decrease of T alleles in MS patients. T allele less frequent in female MS patients Significant increase of T alleles in MS patients
MS: 56/199 (28,1%) Controls: 44/146 (30.1%) MS: 41/187 (21.9%) Controls: 82/282 (29.1%)
Swedish
Zivković et al., 2007
199 cases, 146 controls 187 cases, 282 controls
No significant increase of T alleles in MS patients Significant increase of T alleles in MS patients
(HWE) with p b 10− 6 were excluded. All samples were merged and only SNPs genotyped in all samples were used for further processing. For the imputation, we first prephased the genotype data using SHAPEIT (Delaneau et al., 2011), and subsequently imputed SNPs within the MMP9 gene and the two flanking 2 Mbp regions (hg19:chr20: 42,637,658-46,645,154) to the 1000 genomes references (phase I integrated variant set release (v3) in NCBI build 37 (hg19)) using IMPUTE2 v2.3.0 (Marchini et al., 2007). SNPs with an imputation quality b0.8 (IMPUTE2 info score) and a MAF b1% were excluded. For the final analysis, genotypes of 1421 cases and 1345 controls for 55,884 SNPs were used. On these data, we performed logistic regression using the software tool PLINK v1.07 (Purcell et al., 2007), including age, sex and the first five axes of variation from the MDS analysis as covariates (dosage, allelic model). Only results with an R2 quality metric/information content N 0.9 and b 1.0 were included in downstream analyses (8303 SNPs). Finally, we performed a random-effects meta-analysis on our own imputed and previously published data using R 2.15.2 (http://www.rproject.org/) and the package rmeta. Based on published genotype distributions, ORs and CIs were calculated for the dominant model without covariates. The power of each study was determined using the Genetic Power Calculator (Purcell et al., 2003). 3. Results We performed dosage analyses using an allelic model and including age, sex, and the first five axes of variation from the MDS analysis as covariates. None of these SNPs showed a significant association with MS after correction for multiple testing (Bonferroni adjusted significance level α = 8.95 × 10−7, Fig. 1A). SNP rs6073751 demonstrated the strongest effect with a nominal p-value of 0.00121. This SNP, however, is not located within the MMP9 gene, but in the upstream gene WAP fourdisulfide core domain 2 (WFDC2). At MMP9, the SNP rs3918242 (Fig. 1A,B), which has been previously reported as associated with MS (Table 1), showed a nominally significant effect with a p-value of
Table 2 Demographic data of cases and controls.
Total number Females Males Mean age [years] (+/−standard deviation) Age range
Cases
Controls
1421 70.4% 29.6% 39.3 (+/−10.22) 17–81
1345 62.6% 37.4% 50.0 (+/−13.46) 18–90
Serbian
Brazilian Southern Italian
0.02936 (OR referring to the minor allele (T) = 0.82, CI = 0.69–0.98). This suggests a decreased frequency of the T allele in our MS patients. Separate calculations for females and males revealed a greater reduction of T-alleles in female MS patients in our cohort (OR 1.26 vs 1.16). A meta-analysis using a random-effects model, including six previously published studies and our own imputed data, showed no significant association between SNPs within the MMP9 gene and MS susceptibility when applying a dominant genetic model (Fig. 2). We used a random-effects model, as there was evidence for heterogeneity between the studies (p-value for heterogeneity = 4.7 × 10−7). 4. Discussion The aim of our study was to elucidate the genetic impact of variants within the MMP-9 gene and its two flanking 2 Mbp regions on MS susceptibility. Rs6073751, located within WFDC2, demonstrated the strongest effect. Although it did not remain significant after correction for multiple testing, WFDC2 may be important in the pathophysiology of MS, as it is part of the WAP domain protein family known to modulate immunity (Makarycheva et al., 2011, Wilkinson et al., 2011). Recently, interpopulation heterogeneity of the WFDC gene region has been demonstrated, probably driven by the diversity of pathogens on different continents (Ferreira et al., 2013). This finding is interesting, as MS prevalence decreases from the poles to the equator. Thus, further studies on the regional variants of these genes and their role on autoimmunity are warranted. Subsequently, we focused on rs3918242 (formerly known as − 1562C/T), located close to MMP9, and previously reported as being associated with MS. We found a nominally significant effect towards a decreased T allele frequency in German MS patients, similar to the results by Benesová et al. (2008) observed in a Czech population. However, this finding did not remain significant after correction for multiple testing and is discrepant to results from Southern Italian, Polish, and Brazilian samples, which exhibited an increased T allele frequency in MS patients (Fernandes et al., 2009; Mirowska-Guzel et al., 2009; La Russa et al., 2010). Previous reports have suggested that the T allele is underrepresented only in female MS patients (Zivković et al., 2007 and Benesová et al., 2008). Although the effect on MS susceptibility is not significant, neither in males nor in females, the direction of our results is in line with the data of Benesová et al. and Zivković et al. We performed a meta-analysis on our data together with the data of the studies that had investigated this SNP previously. We found no overall significant effect of this SNP on MS susceptibility. Moreover, based on our German sample, these results do not support common genetic variants in the MMP9 gene and its flanking +/− 2 Mbp region to be risk factors for MS. This is in line with the results of a recently published Australian study and a linkage
48
S. Nischwitz et al. / Journal of Neuroimmunology 279 (2015) 46–49
Fig. 1. A: No SNP within MMP9 or +/−2 Mbp of the gene showed a significant association with MS after correction for multiple testing (Bonferroni adjusted significance level α = 8.95 × 10−7). SNP rs6073751 is labelled in green, rs3918242 in red. B: At the locus MMP9, the SNP rs3918242, which has previously been found associated with MS susceptibility, showed a nominally significant p-value of 0.02936, indicating a decrease of the T-allele in our cohort of German MS patients. However, there was no significant association after correction for multiple testing.
analysis in 100 Russian families (Makarycheva et al., 2011; Cox et al., 2014). Both studies, however, could not be included in the metaanalysis for methodological reasons.
In conclusion, our results do not suggest that MS susceptibility is affected by known genetic variants within MMP9, although MMP9 itself may play an important role in MS pathophysiology.
Fig. 2. Meta-analysis of previously published data and our results on rs3918242 (formerly known as −1562C/T) located upstream to the MMP9 gene, using a dominant model. “Power” indicates the power to detect an association in a dominant model with a nominal p-value of 0.05, using an average minor allele frequency of 0.148 and the ORs indicated for each study (Purcell, 2003).
S. Nischwitz et al. / Journal of Neuroimmunology 279 (2015) 46–49
Statement of conflicts of interests S. Nischwitz received travel grants for the attention of scientific meetings from Merck Serono and TEVA and grant support from Bayer-Schering. P Rieckmann received honoraria for lectures and steering committee meetings from Baxter, Bayer, Biogen Idec, Boehringer-Ingelheim, Bristol-Myers Squibb, Genpharm, Genzyme, Merck KG, Novartis, Pfizer, Roche, Sanofi-Aventis, Siemens, and Teva pharmaceutical Industries. UK Zettl received travel grants for the attention of scientific meetings and honoraria for speaking from Biogen-Idec, Merck-Serono, Bayer, TEVA, Sanofi-Genzyme, Almirall and Novartis. D Buck received compensation for activities with Bayer HealthCare, Biogen Idec, Merck Serono, and Novartis. She was supported by the Commission for Clinical Research of the Faculty of Medicine, Technical University Munich, Abirisk, and the PML Consortium. F Weber received honoraria from Genzyme and Novartis for serving on a scientific advisory board and a travel grant for the attention of a scientific meeting from Merck-Serono, Biogen Idec and Novartis and received grant support from Merck-Serono and the Federal Ministry of Education and Research (BMBF, Projects Biobanking and Omics in ControlMS as part of the Competence Network Multiple Sclerosis). C Wolf, TFM Andlauer, D Czamara, M Ising, T Bettecken and B Mueller-Myhsok report no disclosures.
References Abraham, M., Shapiro, S., Karni, A., Weiner, H.L., Miller, A., 2005. Gelatinases (MMP-2 and MMP-9) are preferentially expressed by Th1 vs. Th2 cells. J. Neuroimmunol. 163, 157–164. Benesová, Y., Vasků, A., Stourac, P., Hladíková, M., Beránek, M., Kadanka, Z., Novotná, H., Bednarík, J., 2008. Matrix metalloproteinase-9 and matrix metalloproteinase-2 gene polymorphisms in multiple sclerosis. J. Neuroimmunol. 205, 105–109. Bernal, F., Elias, B., Hartung, H.P., Kieseier, B.C., 2009. Regulation of matrix metalloproteinases and their inhibitors by interferon-beta: a longitudinal study in multiple sclerosis patients. Mult. Scler. 15, 721–727. Chandler, S., Miller, K.M., Clements, J.M., Lury, J., Corkill, D., Anthony, D.C., Adams, S.E., Gearing, A.J., 1997. Matrix metalloproteinases, tumor necrosis factor and multiple sclerosis: an overview. J. Neuroimmunol. 72, 155–161. Cox, M.B., Bowden, N.A., Scott, R.J., Lechner-Scott, J., 2014. Common genetic variants in the plasminogen activation pathway are not associated with multiple sclerosis. Mult. Scler. 20, 489–491. Delaneau, O., Marchini, J., Zagury, J.F., 2011. A linear complexity phasing method for thousands of genomes. Nat. Methods 9, 179–181.
49
Dressel, A., Mirowska-Guzel, D., Gerlach, C., Weber, F., 2007. Migration of T-cell subsets in multiple sclerosis and the effect of interferon-beta1a. Acta Neurol. Scand. 116, 164–168. Fernandes, K.S., Brum, D.G., Sandrim, V.C., Guerreiro, C.T., Barreira, A.A., Tanus-Santos, J.E., 2009. Matrix metalloproteinase-9 genotypes and haplotypes are associated with multiple sclerosis and with the degree of disability of the disease. J. Neuroimmunol. 214, 128–131. Ferreira, Z., Seixas, S., Andrés, A.M., Kretzschmar, W.W., Mullikin, J.C., Cherukuri, P.F., Cruz, P., Swanson, W.J., Comparative Sequencing Program, N.I.S.C., Clark, A.G., Green, E.D., Hurle, B., 2013. Reproduction and immunity-driven natural selection in the human WFDC locus. Mol. Biol. Evol. 30, 938–950. Kouwenhoven, M., Ozenci, V., Gomes, A., Yarilin, D., Giedraitis, V., Press, R., Link, H., 2001. Multiple sclerosis: elevated expression of matrix metalloproteinases in blood monocytes. J. Autoimmun. 16, 463–470. Kouwenhoven, M., Ozenci, V., Tjernlund, A., Pashenkov, M., Homman, M., Press, R., Link, H., 2002. Monocyte-derived dendritic cells express and secrete matrix-degrading metalloproteinases and their inhibitors and are imbalanced in multiple sclerosis. J. Neuroimmunol. 126, 161–171. La Russa, A., Cittadella, R., De Marco, E.V., Valentino, P., Andreoli, V., Trecroci, F., Latorre, V., Gambardella, A., Quattrone, A., 2010. Single nucleotide polymorphism in the MMP-9 gene is associated with susceptibility to develop multiple sclerosis in an Italian case– control study. J. Neuroimmunol. 225, 175–179. Makarycheva, O.Y., Tsareva, E.Y., Sudomoina, M.A., Kulakova, O.G., Titov, B.V., Bykova, O.V., Gol'tsova, N.V., Kuzenkova, L.M., Boiko, A.N., Favorova, O.O., 2011. Family analysis of linkage and association of HLA-DRB1, CTLA4, TGFB1, IL4, CCR5, RANTES, MMP9 and TIMP1 gene polymorphisms with multiple sclerosis. Acta Nat. 3, 85–92. Marchini, J., Howie, B., Myers, S., McVean, G., Donnelly, P., 2007. A new multipoint method for genome-wide association studies via imputation of genotypes. Nat. Genet. 39, 906–913. Mirowska-Guzel, D., Gromadzka, G., Czlonkowski, A., Czlonkowska, A., 2009. Association of MMP1, MMP3, MMP9, and MMP12 polymorphisms with risk and clinical course of multiple sclerosis in a Polish population. J. Neuroimmunol. 214, 113–117. Nelissen, I., Vandenbroeck, K., Fiten, P., Hillert, J., Olsson, T., Marrosu, M.G., Opdenakker, G., 2000. Polymorphism analysis suggests that the gelatinase B gene is not a susceptibility factor for multiple sclerosis. J. Neuroimmunol. 105, 58–63. Purcell, S., Cherny, S.S., Sham, P.C., 2003. Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics 19, 149–150. Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A., Bender, D., Maller, J., Sklar, P., de Bakker, P.I., Daly, M.J., Sham, P.C., 2007. PLINK: a toolset for whole-genome association and population-based linkage analysis. Am. J. Hum. Genet. 81, 559–575. Sellebjerg, F., Madsen, H.O., Jensen, C.V., Jensen, J., Garred, P., 2000. CCR5 delta32, matrix metalloproteinase-9 and disease activity in multiple sclerosis. J. Neuroimmunol. 102, 98–106. Shiryaev, S.A., Savinov, A.Y., Cieplak, P., Ratnikov, B.I., Motamedchaboki, K., Smith, J.W., Strongin, A.Y., 2009. Matrix metalloproteinase proteolysis of the myelin basic protein isoforms is a source of immunogenic peptides in autoimmune multiple sclerosis. PLoS One 4, e4952. Wilkinson, T.S., Roghanian, A., Simpson, A.J., Sallenave, J.M., 2011. WAP domain proteins as modulators of mucosal immunity. Biochem. Soc. Trans. 39, 1409–1415. Zivković, M., Djurić, T., Dincić, E., Raicević, R., Alavantić, D., Stanković, A., 2007. Matrix metalloproteinase-9 −1562 C/T gene polymorphism in Serbian patients with multiple sclerosis. J. Neuroimmunol. 189, 147–150.
