Identification of acquired somatic mutations in the gene encoding ...

2 downloads 0 Views 585KB Size Report
Jul 13, 2003 - myelodysplasia associated with α-thalassemia (ATMDS), somatic mutations of the gene encoding the chromatin remodeling factor ATRX cause ...
Identification of acquired somatic mutations in the gene encoding chromatin-remodeling factor ATRX in the α-thalassemia myelodysplasia syndrome (ATMDS) Richard J Gibbons1, Andrea Pellagatti2, David Garrick1, William G Wood1, Nicola Malik1, Helena Ayyub1, Cordelia Langford3, Jacqueline Boultwood2, James S Wainscoat2 & Douglas R Higgs1

α-thalassemia is a common inherited form of anemia that usually results from deletion of one or more of the duplicated α-globin genes on chromosome 16 (normal genotype αα/αα; ref. 1). A mild form of α-thalassemia is also associated with a variety of developmental abnormalities in a rare, severe form of X-linked mental retardation (ATR-X syndrome; ref. 2). The gene mutated in this condition (ATRX) encodes a member of the SWI2/SNF2 family of proteins3. Like other members of this group, multiprotein complexes isolated by ATRX antibodies have ATP-dependent nucleosome-remodeling and DNA translocase activities in vitro (Y. Xue et al., manuscript submitted). ATRX is a nuclear protein that localizes to nuclear subcompartments called PML bodies and to pericentromeric heterochromatin, where it interacts with a known component of heterochromatin, HP1 (refs. 4,5 and D.G. et al., manuscript submitted). The consistent clinical and hematological phenotype seen in individuals with ATR-X syndrome

suggest that, as for SWI2/SNF2, there is a discrete group of genes, including α-globin genes, whose expression is perturbed by mutations in ATRX. As for most of these important regulatory proteins, however, the mechanism of its action and its precise role in vivo is unclear. α-thalassemia may rarely occur as an acquired abnormality in individuals with various types of multilineage myelodysplasia (ATMDS syndrome; refs. 6–8). To date, 71 such individuals have been identified, of whom 62 (87%) are males who have a de novo, acquired form of α-thalassemia with hypochromic microcytic anemia (Fig. 1a). A

a

b

c

d 1,000

1.2

β

900

1.1 1.0

800 α/β globin chain biosynthesis rato

Inherited mutations of specific genes have elucidated the normal roles of the proteins they encode by relating specific mutations to particular phenotypes. But many potentially informative mutations in such genes are lethal early in development. Consequently, inherited mutations may not reflect all the functional roles of such proteins. Acquired, somatic defects should reflect a wider spectrum of mutations because they are not prone to negative selection in development. It has been difficult to identify such mutations so far, but microarray analysis provides a new opportunity to do so. Using this approach, we have shown that in individuals with myelodysplasia associated with α-thalassemia (ATMDS), somatic mutations of the gene encoding the chromatin remodeling factor ATRX cause an unexpectedly severe hematological phenotype compared with the wide spectrum of inherited mutations affecting this gene. These findings cast new light on this pleiotropic cofactor, which appears to be an essential component rather than a mere facilitator of globin gene expression.

700 c.p.m.

© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

LETTERS

600

α/β ratio = 0.18

500 400 300 200

α

100 0

0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

0

10

20

30 40 Fraction number

50

60

70

Figure 1 Hematological analysis in individuals with ATMDS. (a) Blood film from individual 1 showing hypochromia, anisocytosis, poikilocytosis and target cells. (b) Blood film of blood from individual 1 stained with brilliant cresyl blue showing the presence of cells with hemoglobin H (β4) inclusions. (c) Globin chain biosynthesis plot showing marked reduction in the ratio of α/β globin synthesis in individual 2. (d) A comparison of the α/β globin chain biosynthesis ratios for individual 2 (open circle) and the mean for normal controls (filled square).

1MRC Molecular Haematology Unit, Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, OX3 9DS UK. 2LRF Molecular Haematology Unit, Nuffield Department of Clinical Laboratory Sciences, John Radcliffe Hospital, Oxford, UK. 3Sanger Centre, Hinxton, Cambridge, UK. Correspondence should be addressed to D.R.H. ([email protected]).

446

VOLUME 34 | NUMBER 4 | AUGUST 2003 NATURE GENETICS

LETTERS Figure 2 Gene expression analysis in an individual with ATMDS. (a) A portion of the microarray: the pseudocolored signals for the two samples (ATMDS aRNA in green and normal aRNA in red) superimposed to show their relative intensity. Yellow spots represent probes binding equivalent target cDNA from both samples; red spots, normal target is more abundant than ATMDS target; green spots, ATMDS target is more abundant than normal target. One of the three ATRX probes is shown. (b) Distribution plot showing the gene expression ratio between granulocytes from individual 1 with ATMDS and granulocytes from a mixed pool of normal individuals. Values in the range 0–0.5 are boxed. (c) Magnification of the distribution plot in the range 0–0.5. The three probes representing ATRX are colored red. (d) Graph comparing real-time quantitative PCR data on ATRX expression in granulocytes from 7 normal controls, 13 individuals with MDS and individual 1 with ATMDS. Values are corrected so that the mean of the normal = 100%.

b

d 160%

140%

120%

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

2.20

2.40

2.60

2.80

3.00

100%

Gene expression ratio

c

80%

60%

40%

20%

0.00

0.10

0.20

0.30

0.40

0%

0.50

© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

a

Control

MDS

ATMDS

reduction in α-globin expression leads to an excess of β-globin chains, which form an abnormal hemoglobin (HbH;β4) that is readily detectable in peripheral blood (Fig. 1b). The α/β mRNA and globin chain synthesis ratios are greatly reduced, and in the most severely affected individuals, α chain synthesis is almost abolished (Fig. 1c,d), implying that all four α genes (αα/αα) are downregulated. This degree of α-thalassemia would be lethal during development if it resulted from an inherited mutation. No structural abnormalities in cis to the α-globin genes have been detected, and the downregulation of α-globin is probably associated with a trans-acting mutation. Clearly ATRX is a plausible candidate for harboring mutations associated with this syndrome, but despite extensive analysis, no structural abnormalities in this gene have yet been identified in individuals with ATMDS. Because MDS is a complex clonal disorder predominantly affecting myeloid rather than lymphoid progenitors, archived samples of unfractionated, mature peripheral blood cells have not provided suitable material for prospective mutational studies of this large (300 kb) gene. Therefore, to obtain a population of cells that should all be affected by an acquired mutation leading to ATMDS, we purified (>95%) granulocytes from the peripheral blood of a newly diagnosed individual. Although ATRX is a candidate, because of its size and our failure to find mutations in previous studies, we chose not to look at selected genes and instead used microarray analysis to look, in a more comprehensive way, for genes whose expression might be perturbed in ATMDS.

NATURE GENETICS VOLUME 34 | NUMBER 4 | AUGUST 2003

We compared total RNA from a male with ATMDS (individual 1) with pooled granulocyte total RNA from seven unaffected controls (four females, three males) using a human cDNA microarray. ATRX was one of the genes with the lowest expression in the ATMDS sample relative to the control (represented by three IMAGE clones on the array; Fig. 2a–c). Only XIST, expressed in females and therefore present in the pooled control but not in the male ATMDS case, was lower. We validated this result using quantitative RT–PCR and found that ATRX expression in granulocytes was 3–4% of that in normal controls (Fig. 2d). In contrast, there was no significant reduction in ATRX expression in a group of 13 individuals with MDS without αthalassemia. Sequence analysis identified a G→A mutation in the canonical splice donor site (GT) of intron 1 (Fig. 3a) of ATRX. This mutation was present in granulocytes but absent in DNA from both buccal cells and a lymphoblastoid cell line derived from individual 1. Because ATMDS is a clonal disorder affecting myeloid progenitors, DNA isolated from the unfractionated white cells (granulocytes and lymphocytes) in peripheral blood had a mixture of mutant and wild-type sequences (Fig. 3b). As purified granulocytes were not available from any other individuals with ATMDS, we used archived bone marrow samples as an enriched rather than a pure source of ATMDS myeloid cells. In each case we systematically sequenced ATRX from bone marrow cDNA. In individual 2, a C→G mutation in exon 4 (Fig. 3c) gives rise to a premature stop (S79X). Again, as expected, we detected a mixture of mutant and wild-type sequence in DNA from bone marrow and peripheral blood; this was confirmed by subcloning the two species (Fig. 3c). In individuals 3 and 4, RT–PCR identified alternative splicing not observed in normal controls (Fig. 4a,b). We subcloned and sequenced the products and found that in individual 3, sequence from intron 18 was spliced into the mRNA and in individual 4, exon 16 was skipped and sequence from intron 16 was included (Fig. 4a,b). The resultant cDNA is predicted to cause truncation of ATRX protein owing to a frameshift. In individuals 3 and 4 the underlying mutations in the genomic DNA have not been identified; this may reflect the difficulty of finding a mutation when it is only present in a proportion of cells in the sample. ATRX seems to be important in regulating gene expression in vivo, and the evidence suggests that it exerts its effects through chromatin. This conclusion is consistent with the different effects of ATRX mutations on α- and β-globin expression, as these genes lie in different chromatin environments9,10 and presumably ATRX is one of many proteins contributing to these regions. From previous observations of ATR-X syndrome, it was clear that the spectrum of 60 different inherited mutations in ATRX alter the balance of α- and β-globin expression, giving rise to a mild form of α-thalassemia2 in which α/β-globin

447

© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

LETTERS Figure 3 ATRX mutation analysis in individuals with ATMDS. (a) Sequence analysis in individual 1 with ATMDS. DNA from buccal cells was compared with that from purified granulocytes. The normal sequence is shown below: exonic sequence is in upper case, intronic sequence is in lower case. The G→A mutation in purified granulocytes is shown. (b) Using modified PCR primers, the mutation in individual 1 with ATMDS destroys an RsaI site. The presence of the restriction site has been used to compare DNA samples from purified granulocytes (lane 1), unfractionated white blood cells from peripheral blood (lane 2), lymphoblastoid cells (lane 3) and buccal cells (lane 4). (c) Sequence analysis in individual 2 with ATMDS. The trace shows the exon 4 sequence in DNA from bone marrow. The normal sequence is shown with the C→G mutation indicated. On the right hand side are shown mutant and wild-type sequences obtained from subclones of DNA cloned from an RT–PCR product derived from bone marrow.

chain synthesis ratios lie in the range 0.3–0.9 (ref. 11 and data not shown). None of these inherited mutations seems to completely abolish ATRX function12. Preliminary studies of ATPase activity in ATRX immunoprecipitates shows residual activity in cell extracts from individuals with ATR-X syndrome (A. Argentaro, MRC Molecular Haematology Unit, Oxford, UK, personal communication). By contrast, Atrx knockout embryos that express no full-length protein do not develop beyond 9.5 d (D.G., unpublished observation). It therefore seems likely that a constitutional null mutation in Atrx is lethal. The severe hematological phenotype observed in individuals with ATMDS results from a tissue-specific ‘knockout’ of ATRX limited to hematological progenitors. This indicates that ATRX does not simply modify the level of α-globin expression but is absolutely required for its expression in vivo. Given the complex yet consistent phenotype of ATR-X syndrome, this chromatin-associated protein must regulate expression of many target genes. By analogy to other chromatin-remodeling complexes, one might expect ATRX to have different effects on the genes it regulates, α-globin representing one extreme. It has been suggested that chromatin-remodeling complexes exert their greatest effect when genes are most tightly repressed, for example, in the wake of mitosis when chromatin is not fully decondensed13. In the case of α-globin, this could be most important when the nucleus condenses during terminal erythroid maturation as globin expression reaches its maximum. ATMDS potentially offers a model within which to address these issues. For many important genes, inherited null mutations are lethal early in development. The only viable manifestation of such mutations in these genes will be seen in diseases associated with acquired somatic mutations. Other examples of this, in addition to ATRX, include mutations of PIGA in paroxysmal nocturnal hemoglobinuria14,15 and GNAS1 in McCune–Albright syndrome16. Identification of this class of acquired somatic mutation using high-throughput screening strategies promises to provide valuable functional information on many other genes in this category.

a

b

c

RNA extraction and pooling. We extracted total RNA using TRIZOL (Invitrogen-Life Technologies) according to the protocol supplied by the manufacturer. Briefly, we resuspended granulocyte pellets in TRIZOL, incubated them at room temperature for 2–3 min, added chloroform to separate the phases and then precipitated RNA with isopropanol from the recovered aqueous phase. We resuspended the RNA pellet in nuclease-free water and quantified it by spectrophotometer measurement. We mixed equal amounts of RNA extracted from the granulocyte fraction from seven healthy volunteers to generate a pool of normal control RNA. RNA labeling and microarray hybridization. In each microarray experiment, we compared RNA from an individual with ATMDS with the normal RNA pool using a T7-based amplification method (MessageAmp aRNA Kit, Ambion). We reverse-transcribed 1 µg of total RNA using a T7-oligo-dT primer and then produced differentially labeled aRNAs by in vitro transcription in the presence of either Cy3-UTP or Cy5-UTP (Amersham Biosciences). We then purified the labeled aRNAs and hybridized them competitively to a microarray slide (Sanger Institute) containing 10,000 spots representing 6,000 known human genes. Hybridization occurred at 47 °C overnight, and we then washed the slides at room temperature once in 2× saline sodium citrate for 5 min, twice in 0.1× saline sodium citrate and 0.1% SDS for 30 min and twice in 0.1× saline sodium citrate for 5 min. The experiment was done in triplicate. Scanning and analysis. We dried the slides by centrifugation, scanned them immediately (ScanArray 4000, Packard BioScience) with a 10-µm resolution and imported the generated TIFF images into QuantArray 3.0 (Packard BioScience). After image and grid alignment and spot location, we normalized the Cy5 and Cy3 intensities to the median value of the whole array and determined the Cy5/Cy3 ratio for every spot. Analysis was done with GeneSpring 4.2 (Silicon Genetics).

a

METHODS Individuals with ATMDS. Ethical approval for this work was obtained from the Oxfordshire Clinical Research Ethics Committee (ref.: C00.196). Informed consent was obtained from all subjects. Figure 4 Abnormal splicing in individuals with ATMDS. Abnormal splicing was identified in individuals 3 and 4 with ATMDS. Horizontal gel analysis of RT–PCR fragments from bone marrow RNA are shown, indicating an additional higher molecular weight fragment in individual 3 (a) and a smaller fragment in individual 4 (b). M, molecular weight marker; 3, individual 3; 4, individual 4; C, control. On the right hand side are depicted the splicing abnormalities. Darkly shaded boxes correspond to exons, lightly shaded boxes to intronic sequence incorporated into the cDNA.

448

b

VOLUME 34 | NUMBER 4 | AUGUST 2003 NATURE GENETICS

© 2003 Nature Publishing Group http://www.nature.com/naturegenetics

LETTERS Real-time quantitative PCR. We reverse-transcribed 2 µg of total RNA from the affected individual and from each of the 7 healthy controls included in the pool and 13 individuals with MDS using RETROscript kit (Ambion). We diluted the generated first-strand cDNA and used it as template for real-time quantitative PCR analysis (TaqMan). Using Primer Express 2.0 software (Perkin Elmer/Applied Biosystems), we designed PCR primers for ATRX in adjacent exons and the fluorogenic probe spanning the junction between them. The fluorogenic probe was custom synthesized by Perkin Elmer/Applied Biosystems; PCR primers were synthesized by MWG Biotech. We used the expression level of the β2-microglobulin gene to normalize for differences in input cDNA. Primer sequences for ATRX are available on request. For the β2microglobulin gene, we used the Human β2M pre-developed TaqMan Assay (Perkin Elmer/Applied Biosystems). PCR reactions occurred in a volume of 25 µl in presence of 7.5 pmol of each primer, 2.5 pmol of fluorogenic probe and 12.5 µl of TaqMan 2× Universal PCR mastermix. We used 5 µl of diluted cDNA solution as template in each reaction. At each cycle of the PCR process, we monitored the increase of the fluorescence by an ABI Prism 5700 Sequence Detection System (Perkin Elmer/Applied Biosystems). The thermal cycling program consisted of a first step at 50 °C for 2 min, then 95 °C for 10 min and finally 40 cycles of amplification at 95 °C for 15 s and 60 °C for 1 min. Each sample was analyzed in triplicate. In each case, we used a reverse transcriptase–negative control to verify the absence of DNA contamination. Mutation analysis. We analyzed ATRX either exon by exon from genomic DNA or by RT–PCR using RNA from bone marrow. For individual 1 with ATMDS, we designed modified primers to create, in the wild-type PCR fragment, an RsaI site that was destroyed by the mutation. Sequencing and PCR primer sequences and protocols are available on request. URL. Microarray details are available at http://www.sanger.ac.uk/Projects/ Microarrays/. GEO accession numbers. Array data was deposited at the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) with accession numbers GSM6788, GSM6789 and GSM6790. ACKNOWLEDGMENTS We thank L. Rose for help preparing the manuscript, A. Argentaro for allowing us to cite his unpublished data, C. Fisher for carrying out the hematology, the referring physicians M. Cook and A. Hendrick, the individuals with ATMDS for their participation, D. Weatherall for bringing this condition to our attention and for his continued support, and previous laboratory members C. Craddock, M. Vickers, C. Hatton and V. Barbour for important contributions to the description of

NATURE GENETICS VOLUME 34 | NUMBER 4 | AUGUST 2003

ATMDS. The work was supported by the Medical Research Council and the Leukaemia Research Fund. COMPETING INTERESTS STATEMENT The authors declare that they have no competing financial interests. Received 4 March; accepted 12 June 2003 Published online 13 July 2003; doi:10.1038/ng1213 1. Higgs, D.R. et al. A review of the molecular genetics of the human α-globin gene cluster. Blood 73, 1081–1104 (1989). 2. Gibbons, R.J. & Higgs, D.R. The molecular-clinical spectrum of the ATR-X syndrome. Am. J. Med. Genet. 97, 204–212 (2000). 3. Gibbons, R.J., Picketts, D.J., Villard, L. & Higgs, D.R. Mutations in a putative global transcriptional regulator cause X-linked mental retardation with α-thalassemia (ATRX syndrome). Cell 80, 837–845 (1995). 4. McDowell, T.L. et al. Localization of a putative transcriptional regulator (ATRX) at pericentromeric heterochromatin and the short arms of acrocentric chromosomes. Proc. Natl. Acad. Sci. USA 96, 13983–13988 (1999). 5. Berube, N.G., Smeenk, C.A. & Picketts, D.J. Cell cycle-dependent phosphorylation of the ATRX protein correlates with changes in nuclear matrix and chromatin association. Hum. Mol. Genet. 9, 539–547 (2000). 6. Weatherall, D.J. et al. Acquired haemoglobin H disease in leukaemia: pathophysiology and molecular basis. Br. J. Haematol. 38, 305–322 (1978). 7. Higgs, D.R., Wood, W.G., Barton, C. & Weatherall, D.J. Clinical features and molecular analysis of acquired HbH disease. Am. J. Med. 75, 181–191 (1983). 8. Higgs, D.R. & Bowden, D.K. Clinical and laboratory features of α-thalassemia syndromes. in Disorders of Hemoglobin (ed. Nagel, R.L.) 431–469 (Cambridge University Press, Cambridge, 2001). 9. Craddock, C.F. et al. Contrasting effects of α and β globin regulatory elements on chromatin structure may be related to their different chromosomal environments. EMBO J. 14, 1718–1726 (1995). 10. Higgs, D.R., Sharpe, J.A. & Wood, W.G. Understanding α-globin gene expression: a step towards effective gene therapy. Semin. Hematol. 35, 93–104 (1998). 11. Wilkie, A.O.M. et al. Clinical features and molecular analysis of the αthalassemia/mental retardation syndromes. II. Cases without detectable abnormality of the α-globin complex. Am. J. Hum. Genet. 46, 1127-1140 (1990). 12. Wada, T. & Gibbons, R.J. ATR-X syndrome. in Genetics and Genomics of Neurobehavioural Disorders (ed. Fisch, G.S.) 309–334 (Humana, Totowa, New Jersey, 2003). 13. Krebs, J.E., Fry, C.J., Samuels, M.L. & Peterson, C.L. Global role for chromatin remodeling enzymes in mitotic gene expression. Cell 102, 587–598 (2000). 14. Takeda, J. et al. Deficiency of the GPI anchor caused by a somatic mutation of the PIG-A gene in paroxysmal nocturnal hemoglobinuria. Cell 73, 703–711 (1993). 15. Nafa, K., Bessler, M., Castro-Malaspina, H., Jhanwar, S. & Luzzatto, L. The spectrum of somatic mutations in the PIG-A gene in paroxysmal nocturnal hemoglobinuria includes large deletions and small duplications. Blood Cells Mol. Dis. 24, 370–384 (1998). 16. Happle, R. The McCune-Albright syndrome: a lethal gene surviving by mosaicism. Clin. Genet. 29, 321–324 (1986).

449