X chromosome inactivation patterns in brain in Rett syndrome: implications for the disease phenotype

Brain & Development 27 (2005) 266 270 Original article X chromosome inactivation patterns in brain in Rett syndrome: implications for the disease phenotype Joanne H. Gibson a,b, Sarah L. Williamson a,b, Susan Arbuckle c, John Christodoulou a,b, * a Metabolic Diseases Research Unit, Western Sydney Genetics Program, Children s Hospital at Westmead, Locked Bag 4001, Westmead, NSW 2145, Australia b School of Paediatrics and Child Health, University of Sydney, Sydney, NSW, Australia c Department of Histopathology, Children s Hospital at Westmead, Sydney, NSW, Australia Received 22 March 2004; received in revised form 23 June 2004; accepted 11 July 2004 www.elsevier.com/locate/braindev Abstract Skewed X chromosome inactivation (XCI) has been implicated in modulating the severity of Rett syndrome (RTT), although studies by different groups have yielded conflicting results. In this study we have characterised the XCI pattern in various neuroanatomical regions of nine RTT brains and non-neural tissue in two of these patients to determine whether or not variable XCI patterns occur in different brain regions or somatic tissues of the same patient. The mean XCI patterns for frontal and occipital were compared between RTT and control subjects, and showed no significant differences when comparing RTT frontal to control frontal or RTT occipital to control occipital. However, one RTT subject displayed variability across the different neuroanatomical regions of the brain and skewing in some non-neural tissues. This observation adds another dimension to the epigenetic factors that may contribute to the phenotype in RTT. It also mandates that caution should be exercised in factoring XCI, including assumptions based on the blood XCI pattern, into the development of phenotype genotype correlations. q 2004 Elsevier B.V. All rights reserved. Keywords: Methyl CpG-binding protein 2; Epigenetics; Disease phenotype; Rett syndrome; X inactivation 1. Introduction Rett syndrome (RTT) is an X linked dominant neurodevelopmental disorder that predominantly affects females. A period of apparently normal development in the first 6 18 months of life is followed by loss of previously acquired skills (such as speech and purposeful hand movements) to varying degrees, and is accompanied by acquired microcephaly and stereotypic hand movements. The phenotypic spectrum is very broad, ranging from the milder speech preserved cases to severe intellectual disability [1]. In up to 80% of clinically diagnosed RTT cases a mutation in the X linked methyl-cpg binding protein 2 Abbreviations: XCI, X chromosome inactivation; RTT, Rett syndrome; AR, androgen receptor; BSA, bovine serum albumin; GTP, guanine triphosphate; MBD, methyl-binding domain; TRD-NLS, transcription repressor domain - nuclear localisation signal. * Corresponding author. Tel.: C61 2 9845 3452; fax: C61 2 9845 1864. E-mail address: johnc@chw.edu.au (J. Christodoulou). (MECP2) gene is identified [2 11]. MECP2 is expressed widely and is particularly abundant in the central nervous system. Not surprisingly then, the clinical presentation of RTT is one of the neurological dysfunction. The RTT brain shows regional pathology, with areas such as the frontal, motor and inferior temporal cortices affected to a greater extent than the occipital (visual), which appears to escape neuropathological changes [12 14]. Skewed X chromosome inactivation (XCI) has been implicated in easing the severity of specific X linked mental retardation disorders due to preferential inactivation of the X chromosome that harbours the mutant allele [15]. Consequently, if the normal allele is primarily expressed this may lead to a lessening of the severity of the disease phenotype. Alternatively, skewed XCI resulting in the X chromosome harbouring the mutant allele being predominantly active, could lead to a more severe disease phenotype. To date, most researchers have examined XCI patterns in peripheral blood DNA with variable results [10,11,16,17]. 0387-7604/$ - see front matter q 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.braindev.2004.07.002

J.H. Gibson et al. / Brain & Development 27 (2005) 266 270 267 There are at least two lines of evidence supporting the hypothesis that skewed X-inactivation might modulate the clinical severity of RTT. Ishii and colleagues reported identical twins with the R294X mutation who had very disparate clinical features, with the more mildly affected twin having skewed XCI, whilst the more severely affected twin had random XCI [18]. Secondly, there have been a number of reports of obligate heterozygotes with only mild or no clinical features of RTT and who had skewed (presumably protective) XCI, whilst their affected daughters had balanced XCI [3,5,19 21]. There have been a small number of studies examining XCI patterns in brain, but they have looked at most one or two brain regions [22 25]. In this study we have characterised the XCI pattern in various neuroanatomical regions of nine RTT brains and non-neural tissue in two patients. This was undertaken in order to examine whether or not variable XCI patterns occur in different brain regions or somatic tissues of the same patient. 2. Methods 2.1. Tissue samples Post mortem neural and somatic tissue was obtained from various sources including The Harvard Brain Tissue Resource Center, Boston, USA, the New South Wales Tissue Resource Centre, NSW, Australia, Baylor College of Medicine, Texas, USA, Telethon Institute for Child Health Research, Perth, Western Australia. Relevant information about the samples is summarised in Table 1. All specimens were retained and used for research with appropriate consent from the families. They were stored at K70 8C from the time of collection as wedges of specified areas and in two cases as cerebral hemisphere, coronally sliced. The outer cortical layers were obtained using a scalpel and forceps whilst the specimen remained frozen on dry ice. 2.2. Mutation analysis Genomic DNA was extracted from the post mortem tissue samples listed in Table 1 using a modified version of Sambrook and Russell [26]. DNA was isolated from the frontal and occipital cortices in each case, and parietal, cerebellum, substantia nigra, thalamus, temporal and hippocampus subject to availability. DNA was also extracted from non-neural tissue from cases RTT 5 (liver, kidney, heart and quadriceps muscle) and RTT 6 (unspecified muscle and peripheral blood). DNA was quantified using a Beckman DU w 650 Spectrophotometer. Mutation analysis was performed using previously described methods [11]. 2.3. X chromosome inactivation assay 2.3.1. Methylation-sensitive X chromosome inactivation assay The methylation status of Hpa II and Hha I sites that flank the highly polymorphic X linked human androgen receptor locus (AR) was assessed in order to determine whether there was skewing of X inactivation at this locus, and by Table 1 Details of the brain samples used in this study Sample Phenotype Age at death (year) Post mortem interval (h) Nucleotide change Amino acid change RTT 1 RTT-TNK 11 9.6 c.763cot R255X HBTRC RTT 2 RTT-TNK 23 10.5 NI NI HBTRC RTT 3 RTT-TNK 12 5 c.808cot R270X HBTRC RTT 4 Classical RTT 18 10 c.473cot T158M Local RTT 5 Classical RTT 11 NK c.316cot R106W Local RTT 6 Classical RTT 21 NK c.808cot R270X Local RTT 7 RTT-TNK 7 NK c.316 COT R106W BCM RTT 8 Classical RTT 8 NK NI NI BCM RTT 9 RTT-TNK 4 NK c.750insc P251 fs BCM C 1 53 24 None None HBTRC C 2 68 21.5 None None HBTRC C 3 58 17.8 None None HBTRC C 4 43 48 None None NSWTRC C 5 42 48 None None NSWTRC C 6 31 49 None None NSWTRC C 7 43 24 None None NSWTRC C 8 46 5 None None NSWTRC C 9 52 31 None None NSWTRC C 10 52 9.5 None None NSWTRC RTT, Rett syndrome; RTT-TNK, Rett syndrome-type not known; C, control. NK, not known; HBTRC, Harvard Brain Tissue Resource Centre; Local, Australian RTT subject; BCM, Baylor College of Medicine, Texas Children s Hospital; NSWTRC, New South Wales Tissue Resource Centre; NI, no mutation identified in the MECP2 coding sequence; Fs, frameshift. Limited phenotypic information was available. In the column labelled Phenotype, RTT refers to a definite diagnosis of RTT syndrome, however more specific information with regards to any variations was not available. Source

268 J.H. Gibson et al. / Brain & Development 27 (2005) 266 270 implication the rest of the genes in cis with the inactivated allele. This assay was carried out using a modified version of a previously described assay [11,27], and each patient sample was assessed in two independent reactions, and the mean of the two used for statistical analyses. 2.3.2. Methylation-dependant X chromosome inactivation assay In cases where the difference of two independent methylation-sensitive X chromosome inactivation assays exceeded 10%, a methylation-dependent X chromosome inactivation assay was also undertaken. This assay utilised the restriction endonuclease McrBC, which cleaves DNA containing methylcytosine, but not unmethylated DNA. Genomic DNA (100 ng) was digested with McrBC (New England BioLabs w Inc.) at 37 8C for 2 h in the presence of 100 mg/ml BSA and 1 mm GTP. Amplification of the androgen receptor locus (for digested and undigested samples) was carried out as for the methylation-sensitive assay. 2.4. Statistical analyses The XCI values from Table 2 represent the mean of two independent assays. These values were determined by the method outlined by Allen et al. [27]. These values did not follow a normal distribution and so a Mann Whitney test was used to compare values for independent samples (RTT frontal with Control frontal ; RTT occipital with Control occipital ) and a Wilcoxon Signed Ranks test for the related samples (RTT frontal with RTT occipital ; Control frontal with Control occipital ). 3. Results 3.1. Mutation analysis Pathogenic mutations were present in the coding region of MECP2 in seven out of nine of the RTT patient samples examined (Table 1). This included two common missense mutations in the methyl-binding domain (MBD) in three patients (c.316cot and c.473cot), two common nonsense mutations in the nuclear localisation signal that lies within the transcription repression domain (TRD-NLS) in three patients (c.763cot and c.808cot) and one rare frameshift mutation in the TRD in one patient (c.750insc). 3.2. General observations XCI patterns were assessed in various neuroanatomical regions of RTT and control post mortem brains. In addition, a limited number of RTT non-neural tissue samples were examined by the same means. Nine out of 10 sets of samples from RTT patients in our possession were heterozygous at the AR locus and were therefore informative for the purpose of this assay. The methylation-sensitive XCI assay was carried out twice for each sample, and the mean percentages displayed in Table 2 represent the average activity of the longer androgen receptor allele in each case. If the two independent assays resulted in values that varied by greater that 10% the methylation-dependant XCI assay was performed. In each case the value obtained was within 10% of one of the previous values obtained using the methylation-sensitive assay and so the mean of these two values was used. Table 2 X chromosome inactivation pattern (expressed as a percentage of the largest androgen receptor allele see methods) in post mortem RTT brain tissues compared to control brain tissues Sample Frontal Occipital Parietal Cerebellum Substantia nigra Thalamus Temporal RTT 1 59 57.5 50.5 RTT 2 68.5 61 RTT 3 33.5 49 RTT 4 49 46.5 50 51 61 55 52 66 RTT 5 31 50 27 54 58 27 24 RTT 6 33 50 49 52 31 50 50 RTT 7 34.5 61.5 56 59.5 RTT 8 59 51 53.5 54 58 40 RTT 9 45 39 36.5 51 65 50 C 1 59 60.5 60 C 2 38 42 78 C 3 38 60 53 C 4 49.5 51 C 5 51 44.5 C 6 57.5 48 C 7 48 53.5 C 8 50 45 C 9 48 54.5 C 10 45 61.5 Hippocampus RTT, Rett syndrome; C, control, Mean XCI of frontal and occipital cortices in RTT and control groups: RTT frontal Z45.8 (G13.9); RTT occipital Z51.7 (G7.2); control frontal Z48.4 (G7.3) and control occipital Z52.1 (G6.7).

J.H. Gibson et al. / Brain & Development 27 (2005) 266 270 269 We defined a skewed XCI pattern to be an average expression of one AR allele being greater than 80%. When the XCI patterns for frontal and occipital were compared between RTT and control subjects, no significant differences were observed: RTT frontal / control frontal (PZ0.486); RTT occipital /control occipital (PZ1.0); RTT frontal / RTT occipital (PZ0.374) and control frontal / control occipital (PZ0.358). 3.3. Individual variability in XCI patterns Balanced XCI patterns were observed in all neuroanatomical regions examined. There was however, significant variability (within the definition of balanced XCI) observed in patient sample RTT 5 (Table 2). This XCI pattern varied from 50:50% in the occipital to 24:76% in the temporal. Non-neural tissue (of the same patient) showed considerable variability in the XCI pattern, with liver having a level of 64:34%, and skeletal muscle 45:55%, whilst skewing was present in heart muscle (20:80%) and kidney (9:91%). RTT 6 had a balanced XCI pattern in both peripheral blood (42:58%), muscle (40:60%) and in all neural tissues examined (Table 2). 4. Discussion Rett syndrome is an X linked disorder with a markedly broad range of clinical severity. The contribution of skewing of XCI in RTT patients has therefore been of considerable interest to researchers as a possible explanation for this variable expressivity. A number of studies have characterised the XCI pattern of RTT patients in peripheral blood samples [10,11,16,17]. However, as RTT manifests itself primarily as a disorder of the central nervous system, it would be of interest to more widely examine the XCI patterns in the regions that are predominantly affected. There have been four other studies examining XCI patterns in RTT brain. Zoghbi et al. [22] found balanced XCI in cortical tissues from three RTT subjects, whilst XCI was moderately skewed in liver, leukocytes and fibroblasts, respectively. Anvret and Wahlström [23] also showed balanced XCI in brain samples from two RTT patients, but did not specify the regions tested. Using a quantitative approach with laser scanning cytometry, LaSalle and colleagues [24] also found balanced XCI in the cerebral and cerebellum of two RTT patients. Finally, Shahbazian and colleagues [25] found random XCI in either cerebral or cerebellum of 10 RTT patients. The latter study included three patients in our study; (RTT1Z B4315, RTT2ZB4321, RTT3ZB4422), however different regions were examined. Our study is the first to characterise the XCI pattern of RTT patients in multiple neuroanatomical regions as well as non-neural tissue. We report that whilst balanced XCI patterns were predominantly observed, individual variability was found across different neuroanatomical regions of one RTT patient brain and skewing was present in the non-neural tissues of this individual. Based on the observations that there are consistent functional [28] and histopathological [14] abnormalities in the brains of RTT patients, we had hypothesised that these regional abnormalities might be due to a consistent pattern of regional skewing of XCI. This study has shown however, that whilst a region of the RTT brain that displays the abovementioned abnormalities (frontal ) consistently has balanced XCI, so does an apparently unaffected region (occipital ). Mild variability exists across different regions of one RTT patient. We were unable to establish whether this is also the case in control brains due to lack of tissue availability. Perhaps more significant is the observation that somatic tissue of the same RTT patient (heart muscle and kidney) had skewed XCI. Therefore, it should not be assumed that the XCI pattern in one particular tissue or cell type reflects the XCI pattern in other tissues or cell types from the same individual. The issue of whether XCI patterns in blood could be used as a reasonable reflection of the pattern in brain has been a vexing one. Cell types that undergo many divisions are more susceptible to skewing due to a growth disadvantage in the cells that harbour the mutated X linked gene. This has been recently reported in lymphoblasts derived from RTT subjects [29]. This suggests that peripheral blood would be more likely to undergo skewing than neural tissue in RTT patients, making the blood XCI pattern an unsuitable surrogate for XCI patterns in brain. In order to determine whether or not XCI does in fact influence the severity of the RTT phenotype, and whether the XCI pattern in blood might be a reasonable surrogate for brain studies, use of the various mouse models for RTT could be of value. One study that accomplished this has recently been reported [30]. Using a mouse model that phenotypically reproduces the human disease, Young and Zoghbi found a skewed pattern of XCI in Purkinje cells in O60% of mutant female mice studied. This skewing exclusively favoured the wildtype allele. Interestingly, two out of seven mice showed varying levels of patterns of XCI in cell populations from different regions of the brain. Using logistic regression, they found that the level of skewing of XCI favouring the wild-type allele was inversely related to the severity of the observed phenotype, suggesting that the phenotype is affected by the XCI pattern even when skewing is not O80%. In summary, our observations of intra-individual variability of XCI across different neural and somatic tissue in at least one RTT patient highlights the importance of exercising caution in factoring blood DNA XCI analysis into the development of phenotype genotype correlations. It is open to conjecture whether regional XCI variation in individual RTT patient brains is consequent upon the mutation itself, or occurs independently of the mutation, but recent evidence in a mouse model [30] would suggest that

270 J.H. Gibson et al. / Brain & Development 27 (2005) 266 270 even hitherto non-significant skewing in brain can have an effect on the clinical phenotype. Acknowledgements We are grateful to the Harvard Brain Tissue Resource Center, Boston, USA; the New South Wales Tissue Resource Centre, Sydney; NSW, Australia; Professor Dawna Armstrong, Baylor College of Medicine, Houston, USA; Dr Anthony Tannenberg, Mater Misericordiae Hospitals, Brisbane, Australia; Dr Ian Andrews, Sydney Children s Hospital, Sydney, NSW, Australia and Dr Helen Leonard, Telethon Institute for Child Health Research, Perth, Australia for provision of the brain samples used in these studies. This research was in funded by the National Health and Medical Research Council of Australia and the Rett Syndrome Australian Research Fund. References [1] Ellaway C, Christodoulou J. Rett syndrome: clinical characteristics and recent genetic advances. Disabil Rehabil 2001;23:98 106. [2] Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY. Rett Syndrome is caused by mutations in X-linked MECP2, encoding methyl-cpg-binding protein 2. Nat Genet 1999; 23:185 8. [3] Wan M, Lee SS, Zhang X, Houwink-Manville I, Song HR, Amir RE, et al. Rett syndrome and beyond: recurrent spontaneous and familial MECP2 mutations at CpG hotspots. Am J Hum Genet 1999;65: 1520 9. [4] Amir RE, Zoghbi HY. Rett syndrome: methyl-cpg-binding protein 2 mutations and phenotype genotype correlations. Am J Med Genet 2000;97:147 52. [5] Bienvenu T, Carrié A, de Roux N, Vinet M-C, Jonveaux P, Couvert P, et al. MECP2 mutations account for most cases of typical forms of Rett syndrome. Hum Mol Genet 2000;9:1377 84. [6] Dragich J, Houwink-Manville I, Schanen C. Rett syndrome: a surprising result of mutation in MECP2. Hum Mol Genet 2000;9: 2365 75. [7] Hampson K, Woods CG, Latif F, Webb T. Mutations in the MECP2 gene in a cohort of girls with Rett syndrome. J Med Genet 2000;37: 610 2. [8] Huppke P, Laccone F, Krämer N, Engel W, Hanefeld F. Rett syndrome: analysis of MECP2 and clinical characterization of 31 patients. Hum Mol Genet 2000;9:1369 75. [9] Obata K, Matsuishi T, Yamashita Y, Fukuda T, Kuwajima K, Horiuchi I, et al. Mutation analysis of the methyl-cpg binding protein 2 gene (MECP2) in patients with Rett syndrome. J Med Genet 2000; 37:608 10. [10] Hoffbuhr KC, Moses LM, Jerdonek MA, Naidu S, Hoffman EP. Associations between MECP2 mutations, X-chromosome inactivation, and phenotype. Ment Retard Dev Disabil Res Rev 2002;8: 99 105. [11] Weaving LS, Williamson SL, Bennetts B, Davis M, Ellaway CJ, Leonard H, et al. Effects of MECP2 mutation type, location and x-inactivation in modulating Rett syndrome phenotype. Am J Med Genet 2003;118A:103 14. [12] Armstrong D, Dunn JK, Antalffy B, Trivedi R. Selective dendritic alterations in the of Rett syndrome. J Neuropathol Exp Neurol 1995;54:195 201. [13] Armstrong DD. Rett syndrome neuropathology review 2000. Brain Dev 2001;23:S72 6. [14] Armstrong DD. Neuropathology of Rett syndrome. Ment Retard Dev Disabil Res Rev 2002;8:72 6. [15] Plenge RM, Stevenson RA, Lubs HA, Schwartz CE, Willard HF. Skewed X-chromosome inactivation is a common feature of X-linked mental retardation disorders. Am J Hum Genet 2002;71:168 73. [16] Amir RE, Van den Veyver IB, Schultz R, Malicki DM, Tran CQ, Dahle EJ, et al. Influence of mutation type and X chromosome inactivation on Rett syndrome phenotypes. Ann Neurol 2000;47: 670 9. [17] Van den Veyver IB, Zoghbi HY. Mutations in the gene encoding methyl-cpg-binding protein 2 cause Rett syndrome. Brain Dev 2001; 23:S147 51. [18] Ishii T, Makita Y, Ogawa A, Amamiya S, Yamamoto M, Miyamoto A, et al. The role of different X-inactivation pattern on the variable clinical phenotype with Rett syndrome. Brain Dev 2001; 23:S161 4. [19] Sirianni N, Naidu S, Pereira J, Pillotto RF, Hoffman EP. Rett syndrome: confirmation of X-linked dominant inheritance, and localization of the gene to Xq28. Am J Hum Genet 1998;63:1552 8. [20] Hoffbuhr K, Devaney JM, LaFleur B, Sirianni N, Scacheri C, Giron J, et al. MeCP2 mutations in children with and without the phenotype of Rett syndrome. Neurology 2001;56:1486 95. [21] Villard L, Lévy N, Xiang F, Kpebe A, Labelle V, Chevillard C, et al. Segregation of a totally skewed pattern of X chromosome inactivation in four familial cases of Rett syndrome without MECP2 mutation: implications for the disease. J Med Genet 2001;38:435 42. [22] Zoghbi HY, Percy AK, Schultz RJ, Fill C. Patterns of X chromosome inactivation in the Rett syndrome. Brain Dev 1990;12:131 5. [23] Anvret M, Wahlström J. Rett syndrome: random X chromosome inactivation. Clin Genet 1994;45:274 5. [24] LaSalle JM, Goldstine J, Balmer D, Greco CM. Quantitative localization of heterogeneous methyl-cpg-binding protein 2 (MeCP2) expression phenotypes in normal and Rett syndrome brain by laser scanning cytometry. Hum Mol Genet 2001;10:1729 40. [25] Shahbazian MD, Sun Y, Zoghbi HY. Balanced X chromosome inactivation patterns in the Rett syndrome brain. Am J Med Genet 2002;111:164 8. [26] Sambrook J, Russell DW., 3rd ed Molecular cloning, a laboratory manual, vol. 1. New York: Cold Spring Harbor Laboratory Press; 2001. [27] Allen RC, Zoghbi HY, Moseley AB, Rosenblatt HM, Belmont JW. Methylation of HpaII and HhaI Sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates with X chromosome inactivation. Am J Hum Genet 1992;51:1229 39. [28] Nomura Y, Segawa M, Hasegawa M. Rett syndrome clinical studies and pathophysiological consideration. Brain Dev 1984;6:475 86. [29] Balmer D, Arredondo J, Samaco RC, LaSalle JM. MECP2 mutations in Rett syndrome adversely affect lymphocyte growth, but do not affect imprinted gene expression in blood or brain. Hum Genet 2002; 110:545 52. [30] Young JI, Zoghbi HY. X-chromosome inactivation patterns are unbalanced and affect the phenotypic outcome in a mouse model of Rett syndrome. Am J Hum Genet 2004;74:511 20.