Abstract

Myelodysplastic syndromes and acute myeloid leukemia with an isodicentric X chromosome [idic(X)(q13)] occur in elderly women and frequently display ringed sideroblasts. Because of the rarity of idic(X)(q13), little is known about its formation, whether a fusion gene is generated, and patterns of additional aberrations. We here present an SNP array study of 14 idic(X)-positive myeloid malignancies, collected through an international collaborative effort. The breakpoints clustered in two regions of segmental duplications and were not in a gene, making dosage effects from the concurrent gain of Xpter-q13 and loss of Xq13-qter, rather than a fusion gene, the most likely pathogenetic outcome. Methylation analysis revealed involvement of the inactive X chromosomes in five cases and of the active in two. The ABCB7 gene, mutated in X-linked sideroblastic anemia and spinocerebellar ataxia, is in the deleted region, suggesting that loss of this gene underlies the frequent presence of ringed sideroblasts. Additional genetic abnormalities were present in 12/14 (86%), including partial uniparental disomies for 9p (one case) and 4q (two cases) associated with homozygous mutations of JAK2 and TET2, respectively. In total, TET2 mutations were seen in 4/11 (36%) analyzed cases, thus constituting a common secondary event in idic(X)-positive malignancies.

INTRODUCTION

An isodicentric X chromosome with breakpoints in Xq13—idic(X)(q13)—is a rare but recurrent abnormality in myeloid malignant disorders, with approximately 30 cases reported in the literature to date (1–16). Notably, all patients with idic(X)-positive neoplasia have been females—presumably because formation of idic(X)(q13) would result in nullisomy for Xq13-qter in males, which would be lethal for the cells—with a median age of 73.5 years at the time of diagnosis (1–16). Most cases are myelodysplastic syndromes (MDS), but a number of acute myeloid leukemia (AML), usually secondary to MDS, has also been reported, as well as a few cases of myeloproliferative neoplasms (MPN) (1–16). Dewald et al. (2), who was the first to describe the idic(X)(q13) as a recurrent aberration in myeloid malignancies, suggested that it may be specifically associated with refractory anemia with ringed sideroblasts (RARS). Although it has subsequently become clear that far from all idic(X)-positive malignancies are of this MDS subtype, pathological iron accumulation is common also in non-RARS cases (8,16). The outcome of idic(X)-positive cases is variable; some investigators report aggressive and rapidly fatal disease and others a relatively favorable clinical course, with survival for several years despite the generally advanced age of the patients (2,5,6,8,16).

In myeloid disorders with idic(X)(q13), this is frequently the only cytogenetic abnormality, albeit with one to three extra copies of the isodicentric chromosome being present in approximately half of the cases (1–16). One normal X chromosome is always retained. The exact breakpoints in Xq13 have never been defined, but two previous fluorescence in situ hybridization studies showed that a ∼400 kb region between the markers DXS1164 and DXS227 in Xq13 was targeted in the total of four MDS cases investigated (13,14). This region is enriched for repeated sequences, in particular LINE repeats, which may facilitate the formation of an isodicentric chromosome (13,14), and contains several genes. However, it is as yet unknown whether any of these are specifically targeted. Thus, the pathogenetic outcome of the idic(X)(q13) remains unknown; formation of a fusion gene, deregulation of a gene close to the breakpoint or dosage effects resulting from the concurrent gain of Xpter-q13 and loss of Xq13-qter are equally possible. The fact that idic(X)(q13) often occurs as the sole cytogenetic abnormality suggests that it may in itself be sufficient for leukemogenesis, although additional submicroscopic aberrations, arising prior to or after the isodicentric chromosome, cannot be excluded.

It is also not known whether the idic(X)(q13) arises via a translocation between two X chromosomes in a cell with trisomy X, between two different X chromatids or via sister chromatid fusion during the G2 phase of the cell cycle. Furthermore, it is unclear whether the abnormality involves the active or the inactive X chromosome (Xa and Xi, respectively). Dewald et al. (16) reported that the idic(X)(q13) appeared to be early-replicating in two cases and thus probably active, but only a few cells could be investigated. In contrast, Temperani et al. (6) found that the isodicentric X chromosome was late-replicating in both of their cases, suggesting involvement of the Xi. Also Rack et al. (14) detected involvement of the Xi, as evidenced by it being late-replicating as well as by methylation of the MAOA locus at Xp11.3 and expression levels of genes on the X chromosome. Thus, although results have been conflicting, available data seem to favor preferred involvement of the Xi in idic(X)(q13).

In the present study, we have investigated a large cohort of idic(X)-positive myeloid malignancies, collected through an international collaborative effort. The aims of this study were 2-fold: firstly, to fine-map the breakpoints of the idic(X)(q13) to determine whether the aberration targets a specific gene in this region, and secondly, to identify any additional genetic abnormalities that may be of pathogenetic importance in these cases. Using SNP array and molecular analyses, we show that the breakpoints in idic(X)-positive cases cluster in two different regions of Xq13 that do not contain any genes, that either of the active and the inactive X chromosomes may be involved, and that TET2 mutations are frequently present.

RESULTS

SNP array analysis reveals two breakpoint clusters in idic(X)-positive cases

The idic(X)(q13) was visible as gain of Xpter-q13 and hemizygous loss of Xq13-qter in 11 of the 14 investigated cases (Figs 1 and 2; Table 2). The breakpoints clustered in two different regions in Xq13, at ∼70.8–70.9 Mb (five cases) and ∼72.1–72.3 Mb (six cases) (Fig. 2). In none of the isodicentric chromosomes was the breakpoint in a gene. In case 6, two microdeletions were seen adjacent to the idic(X)(q13) breakpoint (Table 3). For cases 5, 7 and 8, no copy number abnormalities could be seen on the X chromosome. For case 5, this could be due to normal cell contamination, whereas for cases 7 and 8, the idic(X)(q13) was only present in subclones (Table 1).

Figure 1.

SNP array data showing the X chromosomes in three cases. Top panels show log2 ratios along the X chromosome. Each dot represents the log2 ratio of one marker. A log2 ratio of zero corresponds to a normal, diploid copy number. Increased and decreased log2 ratios correspond to gained and deleted regions, respectively. Lower panels show B allele frequencies, which are calculated as signal intensity for allele B/(signal intensities for allele A + allele B). Homozygous SNPs have a value of 0 or 1 and heterozygous SNPs a value of 0.5. (A) Case 4:1. A shift in log2 ratio is seen at Xq13, representing the concurrent gain of Xpter-q13 and hemizygous loss of Xq13-qter. The B allele frequencies change from values of ∼0.33 and ∼0.67, corresponding to three copies of Xpter-q13, to 0 and 1, corresponding to 1 copy of Xq13-qter, at the Xq13 breakpoint. These B allele frequency values show that there was a minimal contamination of normal cells in the sample. (B) Case 12. As above, shifts in the log2 ratio and B allele frequencies are seen at Xq13, corresponding to the idic(X)(q13) breakpoint. The B allele frequency values deviate from those observed in (A) because of normal cell contamination. (C) Case 11. Increased log2 ratios are seen in the gained region of X compared with (A) and (B) because of the extra copy of the isodicentric chromosome (Table 1). This is also visible in the B allele frequency values of ∼0.25 and ∼0.75, respectively. The panels were extracted from the BeadStudio 3.1.3.0 software with Illumina Genome Viewer 3.2.9 (Illumina).

Figure 1.

SNP array data showing the X chromosomes in three cases. Top panels show log2 ratios along the X chromosome. Each dot represents the log2 ratio of one marker. A log2 ratio of zero corresponds to a normal, diploid copy number. Increased and decreased log2 ratios correspond to gained and deleted regions, respectively. Lower panels show B allele frequencies, which are calculated as signal intensity for allele B/(signal intensities for allele A + allele B). Homozygous SNPs have a value of 0 or 1 and heterozygous SNPs a value of 0.5. (A) Case 4:1. A shift in log2 ratio is seen at Xq13, representing the concurrent gain of Xpter-q13 and hemizygous loss of Xq13-qter. The B allele frequencies change from values of ∼0.33 and ∼0.67, corresponding to three copies of Xpter-q13, to 0 and 1, corresponding to 1 copy of Xq13-qter, at the Xq13 breakpoint. These B allele frequency values show that there was a minimal contamination of normal cells in the sample. (B) Case 12. As above, shifts in the log2 ratio and B allele frequencies are seen at Xq13, corresponding to the idic(X)(q13) breakpoint. The B allele frequency values deviate from those observed in (A) because of normal cell contamination. (C) Case 11. Increased log2 ratios are seen in the gained region of X compared with (A) and (B) because of the extra copy of the isodicentric chromosome (Table 1). This is also visible in the B allele frequency values of ∼0.25 and ∼0.75, respectively. The panels were extracted from the BeadStudio 3.1.3.0 software with Illumina Genome Viewer 3.2.9 (Illumina).

Figure 2.

The Xq13 breakpoints cluster at ∼70.9 and ∼72.1 Mb. Positions are shown in Mb along the X chromosome according to NCBI Build 36. The 11 informative cases are shown at the top, with horizontal lines representing individual breakpoint regions. As seen, several cases (#6/#10, #1/#14 and #4/#11/#12) displayed identical breakpoints as determined by SNP array analysis. Genes are represented by filled boxes and the two regions of segmental duplications are shown in grey. The gene positions were extracted from the Ensembl Genome browser (www.ensembl.org) and the segmental duplications from the UCSC Genome Browser, Human Chained Self Alignments track (http://genome.ucsc.edu/cgi-bin/hgTrackUi?g=chainSelf&hgsid=139789829).

Figure 2.

The Xq13 breakpoints cluster at ∼70.9 and ∼72.1 Mb. Positions are shown in Mb along the X chromosome according to NCBI Build 36. The 11 informative cases are shown at the top, with horizontal lines representing individual breakpoint regions. As seen, several cases (#6/#10, #1/#14 and #4/#11/#12) displayed identical breakpoints as determined by SNP array analysis. Genes are represented by filled boxes and the two regions of segmental duplications are shown in grey. The gene positions were extracted from the Ensembl Genome browser (www.ensembl.org) and the segmental duplications from the UCSC Genome Browser, Human Chained Self Alignments track (http://genome.ucsc.edu/cgi-bin/hgTrackUi?g=chainSelf&hgsid=139789829).

Table 1.

Clinical and cytogenetic features of the idic(X)-positive hematologic cases

Case no. Sex/age Diagnosis Karyotype 
F/60 AML 47,X,idic(X)(q13),+idic(X)(q13) 
F/62 AML 46,X,idic(X)(q13) 
F/67 AML 47,X,idic(X)(q13),+idic(X)(q13) 
4:1 F/55 AML 47,X,idic(X)(q13),+21 
4:2 /56 AML 47,X,idic(X)(q13),+21/46,idem-7 
F/71 AML 46,X,idic(X)(q13) 
F/70 AML 46,X,idic(X)(q13) 
F/66 MDS 46,XX,del(20)(q11.2q13.3)/46-47,X,idic(X)(q13),+idic(X)(q13) 
F/63 MDS 46,XX,del(5)(q13q33)/45,idem,idic(X)(q13),-7 
F/69 RARS 46,X,idic(X)(q13) 
10 F/87 RARS 46,X,idic(X)(q13) 
11 F/75 RAEB 47,X,idic(X)(q13),+del(1)(p11p35)/48,idem + idic(X)(q13) 
12 F/67 RAEB-1 46,X,idic(X)(q13) 
13 F/82 RAEB-2 46,X,idic(X)(q13) 
14 F/78 CMML 46,X,idic(X)(q13) 
Case no. Sex/age Diagnosis Karyotype 
F/60 AML 47,X,idic(X)(q13),+idic(X)(q13) 
F/62 AML 46,X,idic(X)(q13) 
F/67 AML 47,X,idic(X)(q13),+idic(X)(q13) 
4:1 F/55 AML 47,X,idic(X)(q13),+21 
4:2 /56 AML 47,X,idic(X)(q13),+21/46,idem-7 
F/71 AML 46,X,idic(X)(q13) 
F/70 AML 46,X,idic(X)(q13) 
F/66 MDS 46,XX,del(20)(q11.2q13.3)/46-47,X,idic(X)(q13),+idic(X)(q13) 
F/63 MDS 46,XX,del(5)(q13q33)/45,idem,idic(X)(q13),-7 
F/69 RARS 46,X,idic(X)(q13) 
10 F/87 RARS 46,X,idic(X)(q13) 
11 F/75 RAEB 47,X,idic(X)(q13),+del(1)(p11p35)/48,idem + idic(X)(q13) 
12 F/67 RAEB-1 46,X,idic(X)(q13) 
13 F/82 RAEB-2 46,X,idic(X)(q13) 
14 F/78 CMML 46,X,idic(X)(q13) 

F, female; AML, acute myeloid leukemia; MDS, myelodysplastic syndrome; RARS, refractory anemia with ringed sideroblasts; RAEB, refractory anemia with excess blasts; CMML, chronic myelomonocytic leukemia.

Table 2.

The Xq13 breakpoints in the 11 informative idic(X)-positive myeloid malignancies as determined by SNP array analysis

Case no. Breakpoint regiona (base pair position on the X chromosome) 
13 70,873,926–70,898,769 
6b 70,922,466–70,940,409 
10 70,922,446–70,940,409 
70,940,409–70,941,205 
14 70,940,409–70,941,205 
72,113,887–72,116,388 
72,116,388–72,119,355 
11 72,116,388–72,119,355 
12 72,116,388–72,119,355 
72,116,388–72,124,144 
72,265,440–72,279,106 
Case no. Breakpoint regiona (base pair position on the X chromosome) 
13 70,873,926–70,898,769 
6b 70,922,466–70,940,409 
10 70,922,446–70,940,409 
70,940,409–70,941,205 
14 70,940,409–70,941,205 
72,113,887–72,116,388 
72,116,388–72,119,355 
11 72,116,388–72,119,355 
12 72,116,388–72,119,355 
72,116,388–72,124,144 
72,265,440–72,279,106 

aBreakpoint region is given as the position of the last retained SNP on the idic(X) and the position of the first deleted SNP. All positions are according to NCBI Build 36.

bAssociated with gains and losses close to the breakpoint, see Table 3.

Table 3.

Additional genetic abnormalities in the 14 idic(X)-positive cases investigated with SNP array analysis

Case no. Abnormality Chr Start position (bp)a End position (bp)a Size (max. size)b Comment 
HeL 97,453,930 (97,453,308) 97,462,324 (97,465,911) 8,394 (12,603)  
pUPD 175,173,422 (175,144,970) qter 5,751,838 (5,770,290)  
4:1 Gain 21 pter qter 48,129,895 Trisomy 21 
4:2 HeL pter qter 159,138,663 Monosomy 7 
4:2 Gain 21 pter qter 48,129,895 Trisomy 21 
HeL 126,533,570 (126,531,252) 129,785,490 (129,794,607) 3,251,920 (3,263,355)  
HeL 94,171,374 (94,170,331) 94,360,941 (94,377,341) 189,567 (207,010)  
HeL 16 75,874,288 (75,873,524) 75,924,161 (75,927,639) 49,873 (54,115)  
HeL 61,599,263 (58,574,337) 65,300,888 (65,303,992) 3,701,625 (6,729655)  
HeLc 65,303,992 (65,300,888) 67,320,448 (67,530,820) 2,016,456 (2,229,932)  
HeL 20 30,313,178 (30,272,877) 49,236,052 (49,373,633) 18,922,874 (19,100,756) del(20)(q11.21q13.13) 
HeL 82,710,135 (82,583,322) 166,837,468 (166,911,476) 84,127,333 (84,328,154) del(5)(q14.2q34) 
pUPD pter 24,975,964 (25,271,912) 24,975,964 (25,271,912) Associated with JAK2 mutation 
pUPD 137,575,713 (137,552,586) qter 33,539,354 (33,562,481)  
10 pUPD 71,769,844 (71,765,873) qter 119,384,432 (119,388,403) Associated with TET2 mutation 
11 HeL 86,798,872 (86,791,897) 86,824,512 (86,826,618) 25,640 (34,721)  
11 Gain cen qter 124,250,620  
13 Gain 12 77,061,069 (77,058,364) 77,142,694 (77,151,146) 81,625 (92,782)  
14 pUPD 79,108,827 (78,859,722) qter 112,045,449 (112,294,554) Associated with TET2 mutation 
Case no. Abnormality Chr Start position (bp)a End position (bp)a Size (max. size)b Comment 
HeL 97,453,930 (97,453,308) 97,462,324 (97,465,911) 8,394 (12,603)  
pUPD 175,173,422 (175,144,970) qter 5,751,838 (5,770,290)  
4:1 Gain 21 pter qter 48,129,895 Trisomy 21 
4:2 HeL pter qter 159,138,663 Monosomy 7 
4:2 Gain 21 pter qter 48,129,895 Trisomy 21 
HeL 126,533,570 (126,531,252) 129,785,490 (129,794,607) 3,251,920 (3,263,355)  
HeL 94,171,374 (94,170,331) 94,360,941 (94,377,341) 189,567 (207,010)  
HeL 16 75,874,288 (75,873,524) 75,924,161 (75,927,639) 49,873 (54,115)  
HeL 61,599,263 (58,574,337) 65,300,888 (65,303,992) 3,701,625 (6,729655)  
HeLc 65,303,992 (65,300,888) 67,320,448 (67,530,820) 2,016,456 (2,229,932)  
HeL 20 30,313,178 (30,272,877) 49,236,052 (49,373,633) 18,922,874 (19,100,756) del(20)(q11.21q13.13) 
HeL 82,710,135 (82,583,322) 166,837,468 (166,911,476) 84,127,333 (84,328,154) del(5)(q14.2q34) 
pUPD pter 24,975,964 (25,271,912) 24,975,964 (25,271,912) Associated with JAK2 mutation 
pUPD 137,575,713 (137,552,586) qter 33,539,354 (33,562,481)  
10 pUPD 71,769,844 (71,765,873) qter 119,384,432 (119,388,403) Associated with TET2 mutation 
11 HeL 86,798,872 (86,791,897) 86,824,512 (86,826,618) 25,640 (34,721)  
11 Gain cen qter 124,250,620  
13 Gain 12 77,061,069 (77,058,364) 77,142,694 (77,151,146) 81,625 (92,782)  
14 pUPD 79,108,827 (78,859,722) qter 112,045,449 (112,294,554) Associated with TET2 mutation 

Chr, chromosome; bp, base pair; HeL, hemizygous loss; cen, centromere; qter, terminal of q arm; pUPD, partial uniparental disomy; pter, terminal of p arm.

aPositions are given for first abnormal SNP and, in brackets, for first normal neighboring SNP. All positions are according to NCBI Build 36.

bDistance between first and last abnormal SNPs and, in brackets, between first and last normal neighboring SNPs.

cTwo copies remaining in a trisomic region.

SNP array analysis detects additional genetic abnormalities in idic(X)-positive cases

In addition to identifying the idic(X)(q13) breakpoints, SNP array analysis revealed a total of 18 other genetic abnormalities in 12 of the 14 cases (Table 3). These comprised 10 hemizygous deletions, three gains and five pUPDs. The median number of additional changes per case was one (range 0–5). The only aberration found in more than one case was pUPD for 4q21.21-qter, seen in cases 10 and 14.

Acquired pUPDs are associated with JAK2 and TET2 mutations in idic(X)-positive cases

The SNP array analysis revealed pUPD for ∼25 Mb on 9p, including the JAK2 locus at 9p24, in case 8 (Table 3). Analysis of JAK2 showed homozygosity for the p.Val617Phe mutation. None of the other investigated cases carried this mutation. Furthermore, the SNP array analysis showed that cases 10 and 14 both harbored pUPD for 4q, including TET2 at 4q24 (Table 3). Mutation analysis of TET2 showed that both of these cases carried homozygous mutations leading to a stop codon and an amino acid change in a conserved region of the gene, respectively (Table 4). In addition, cases 3 and 13 had heterozygous frame shift mutations in TET2 (Table 4). Thus, TET2 mutations were detected in 4 (36%) of the 11 analyzed cases.

Table 4.

TET2 mutations in idic(X)-positive myeloid malignancies

Case no. DNA mutation Protein mutation Heterozygous/ homozygous pUPD for 4q24 
c.1037_1038insC Frame shift Heterozygous No 
10 c.2016C > T p.Arg544X Homozygous Yes 
13 3153_3154insCATA Frame shift Heterozygous No 
14 c.4495G > A p.Gly1370Glu Homozygous Yes 
Case no. DNA mutation Protein mutation Heterozygous/ homozygous pUPD for 4q24 
c.1037_1038insC Frame shift Heterozygous No 
10 c.2016C > T p.Arg544X Homozygous Yes 
13 3153_3154insCATA Frame shift Heterozygous No 
14 c.4495G > A p.Gly1370Glu Homozygous Yes 

pUPD, partial uniparental disomy.

Either of the active and the inactive X chromosomes may be involved in the idic(X)(q13)

To determine whether the active or the inactive X was involved in idic(X)(q13), HUMARA analysis was performed, investigating the methylation status of the AR gene in Xq12 in the gained region of the isodicentric chromosome. Initially, semi-quantitative PCR was performed on genomic DNA to identify the microsatellite allele corresponding to the X chromosome involved in the idic(X)(q13) in each case (Fig. 3). Subsequently, genomic DNA was digested with HpaII and again used for semi-quantitative PCR. Because the sequence recognized by HpaII is protected by a methyl group on the inactive X chromosome, the digestion pattern revealed whether the idic(X)(q13) was active or not (Fig. 3). Of the 12 cases studied, five were uninformative due to normal cell contamination (four cases) or homozygosity (one case). Of the seven informative cases, the Xi was involved in five (71%) cases (#1, #3, #11, #13 and #14) and the Xa in two (29%) cases (#6 and #8).

Figure 3.

HUMARA analysis shows whether the idic(X)(q13) is active or not. The panels show size-separation of PCR products by capillary electrophoresis. The X axes show the sizes in base pairs and the Y axes the fluorescence activity. Each peak corresponds to one allele of the amplified microsatellite. Top panels: a polymorphic region from the human androgen receptor gene was amplified by semi-quantitative PCR. The amplified region in Xq12 is present in two copies in the isodicentric X chromosome, detectable as an increased peak height. Lower panels: during digestion with the methylation-specific restriction enzyme HpaII, alleles present on the active X chromosome will be digested whereas alleles present on the inactive X will be protected. (A) Case 11. Two alleles are visible, at 239 and 259 base pairs. Semi-quantitative PCR (top panel) showed that the 259-base pairs allele was located on the idic(X)(q13), visible as a peak height ratio of 2.1. After digestion, only the 259-base pair peak was present (lower panel), showing that the isodicentric chromosome was inactive. (B) Case 6. Two alleles are visible, at 197 and 239 base pairs. Semi-quantitative PCR (top panel) showed that the 239-base pairs allele was located on the idic(X)(q13), visible as a peak height ratio of 1.6. After digestion, only the 197-base pair peak was present (lower panel), showing that the isodicentric chromosome was active.

Figure 3.

HUMARA analysis shows whether the idic(X)(q13) is active or not. The panels show size-separation of PCR products by capillary electrophoresis. The X axes show the sizes in base pairs and the Y axes the fluorescence activity. Each peak corresponds to one allele of the amplified microsatellite. Top panels: a polymorphic region from the human androgen receptor gene was amplified by semi-quantitative PCR. The amplified region in Xq12 is present in two copies in the isodicentric X chromosome, detectable as an increased peak height. Lower panels: during digestion with the methylation-specific restriction enzyme HpaII, alleles present on the active X chromosome will be digested whereas alleles present on the inactive X will be protected. (A) Case 11. Two alleles are visible, at 239 and 259 base pairs. Semi-quantitative PCR (top panel) showed that the 259-base pairs allele was located on the idic(X)(q13), visible as a peak height ratio of 2.1. After digestion, only the 259-base pair peak was present (lower panel), showing that the isodicentric chromosome was inactive. (B) Case 6. Two alleles are visible, at 197 and 239 base pairs. Semi-quantitative PCR (top panel) showed that the 239-base pairs allele was located on the idic(X)(q13), visible as a peak height ratio of 1.6. After digestion, only the 197-base pair peak was present (lower panel), showing that the isodicentric chromosome was active.

DISCUSSION

The breakpoints of the isodicentric X chromosome in myeloid malignancies have previously been mapped to an approximately 400 kb region, containing several possible target genes, in Xq13 (13,14). We investigated 11 cases with high-resolution SNP array analysis and found evidence for two distinct breakpoint clusters, at approximately 70.9 Mb (five cases) and 72.1 Mb (seven cases) on the X chromosome (Fig. 2). The latter of these was in the previously reported 400 kb region (13,14), which we could narrow down to 10 kb that encompassed all breakpoints except one (Table 2). However, none of the 11 breakpoints occurred in a gene, strongly indicating that the idic(X)(q13) does not result in the formation of a fusion gene. Instead, the functional outcome of the abnormality is more likely gene dosage effects due to the concurrent gain of Xpter-q13 and loss of Xq13-qter.

Interestingly, both Xq13 breakpoint clusters were located in one of two large segmental duplications (UCSC Genome Browser, Human Chained Self Alignments track; http://genome.ucsc.edu/cgi-bin/hgTrackUi?g=chainSelf&hgsid=139789829) (Fig. 2). Each of these segmental duplications contains inverted sequences with very high sequence similarities. It is highly likely that these repeats contributed to the formation of the idic(X)(q13), perhaps by facilitating sister chromatid exchange, similar to the underlying mechanism for isochromosome 17q—or rather idic (17)(p11)—in various hematologic malignancies and medulloblastomas (17–19).

Based on the fact that idic(X)-positive myeloid disorders frequently display pathological ringed sideroblasts, Dewald et al. (16) speculated that the gene targeted by the idic(X)(q13) may be the same that is involved in X-linked sideroblastic anemia and spinocerebellar ataxia. The gene mutated in this disease has since been identified as ABCB7, which encodes a mitochondrial ATP-binding cassette iron transporter (20). Interestingly, ABCB7 is located at ∼74.3 Mb in Xq13 and is hence always hemizygously deleted in idic(X)-positive cases. Recently, Boultwood et al. (21) showed that ABCB7 is underexpressed in RARS compared with other MDS subgroups and that decreasing expression levels were correlated with increasing percentages of ringed sideroblasts. Taking this into account, we suggest that the hemizygous deletion of ABCB7 resulting from the isodicentric chromosome is the underlying cause of the frequent presence of pathological ringed sideroblasts in myeloid malignancies with idic(X), in line with Dewald et al.'s (16) early hypothesis. This may particularly be true in cases where the isodicentric chromosome involves the Xa.

Considering that the most likely pathogenetic outcome of the idic(X)(q13) is dosage effects, the activation status of the involved X chromosome becomes important. Previous data have favored preferential involvement of the Xi in the idic(X)(q13) (6,14,16). In the present study, the isodicentric chromosome was inactive in five cases and active in two. Hence, the idic(X)(q13) appears to be leukemogenic irrespective of Xa or Xi involvement. Whether there are clinical or morphological differences between these two groups is an interesting issue that remains to be addressed. For example, loss of ABCB7 may result in pathological iron accumulation only if the remaining normal X chromosome is inactive. Unfortunately, we could not test this hypothesis due to lack of clinical information in the two cases with confirmed involvement of the Xa in the idic(X); no data on the possible presence of ringed sideroblasts were available. On the other hand, it is also possible that the pathogenetically important loci affected by the idic(X)(q13) are not subject to X inactivation, such as genes in the pseudoautosomal regions, or that the formation of the isodicentric chromosome leads to deregulation of the inactivation mechanism. In the latter scenario, it is noteworthy that the breakpoints in the idic(X)(q13) are 0.7–2.2 Mb proximal of the XIST gene, which induces the inactivation of one of the X chromosomes. Hence, XIST is deleted. However, XIST expression is generally not needed for maintenance of X inactivation once it has been established (22).

The SNP array analysis revealed chromosomal aberrations in addition to the idic(X)(q13) in most cases (79%) (Table 3). The majority were hemizygous deletions, of which none were recurrent. Several regions of pUPDs were detected, in line with previous SNP array studies of AML, MDS and MPN (23,24). Interestingly, two cases harbored pUPD of 4q. Recently, the tumor suppressor gene TET2, located at 4q24, has been shown to be mutated in a large proportion of myeloid malignancies, sometimes in association with UPD for 4q (25–27). We therefore performed a mutation screening of this gene and found homozygous mutations in both cases with pUPD4q and heterozygous mutations in two additional cases, giving a total TET2 mutation frequency of 36% (4/11 cases; Table 4). Thus, inactivating mutations of TET2 is a common event in idic(X)-positive disorders. The mutations were present in one AML, two MDS and one chronic myelomonocytic leukemia (CMML), showing that clinically and morphologically different idic(X)-positive malignancies have common genetic features in addition to the isodicentric X chromosome.

In conclusion, we here show that the Xq13 breakpoints in myeloid disorders with idic(X)(q13) cluster in two close but separate regions and that they are not in a gene. The most likely pathogenetic outcome is hence gene dosage effects, possibly including the hemizygous deletion of ABCB7, which is associated with X-linked sideroblastic anemia and spinocerebellar ataxia. The vast majority of cases (12/14; 86%) harbor additional genetic abnormalities, most frequently copy number changes and mutations in TET2.

MATERIALS AND METHODS

Patients

The study included 14 patients with an idic(X)-positive myeloid malignant disorder (Table 1). All were females with a median age of 68 years at diagnosis (range 55–87 years). Six cases were AML, two were unclassified MDS, two were RARS, three were refractory anemia with excess blasts (RAEB) and one was a CMML. In 10 of the cases, the idic(X)(q13) was the only cytogenetic aberration (present in one or two copies). The only other recurrent abnormality was monosomy 7 (#4:2 and #8; Table 1). The study was reviewed and approved by the research ethics board at Lund University, and informed consent was obtained in accordance with the Declaration of Helsinki.

SNP array analysis

DNA was extracted according to standard methods from bone marrow samples acquired at diagnosis. For case 4, DNA was also extracted from a sample obtained 1 year after initial diagnosis. For SNP array analysis, the Illumina 1 M-duo bead Infinium BD BeadChip platform was used, containing 1.2 million markers with a median physical distance between markers of 1.5 kb (Illumina, San Diego, CA). SNP array analysis was done according to the manufacturer's instructions (Illumina) and data analysis was performed using the BeadStudio 3.1.3.0 software with Illumina Genome Viewer 3.2.9 (Illumina). Constitutional copy number polymorphisms were excluded based on comparison with the Database of Genomic Variants (http://projects.tcag.ca/variation/) (28).

HUMARA assays

To investigate the methylation status of the isodicentric X chromosome, human androgen receptor gene AR (HUMARA) assays (29) were performed in all cases except #5 and #9, where no further material was available. To ascertain the methylation status, samples were digested with HpaII before PCR amplification of the first exon of the AR gene, located in Xq12, using the forward primer 5′-HEX-CCGAGGAGCTTTCCAGAATC-3′ and the reverse primer 5′-TACGATGGGCTTGGGGAGAA-3′. Undigested DNA was amplified using the same primers to detect gain of Xq12. All PCR products were size-separated using fragment analysis.

Mutation analyses of JAK2 and TET2

Screening for the p.Val617Phe mutation of JAK2 was performed with allele-specific PCR according to Baxter et al. (30) in all cases except #9, for which no material was available. For TET2 mutation status, all coding exons were analyzed by direct sequencing (25) in all cases except #5, #6 and #9, which were excluded due to lack of sufficient amounts of DNA. Sequence analysis and identification of mutations were done using the Mutation Surveyor software, version 3.23 (SoftGenetics, State College, PA).

FUNDING

This study was supported by grants from the Swedish Cancer Society and the Swedish Research Council.

ACKNOWLEDGEMENTS

We are grateful to Dr Lucienne Michaux for help with identifying cases and to Dr Olivier Bernard for providing TET2 primer sequences. The SNP array experiments were performed by Sciblu Genomics at Lund University, Lund, Sweden. The authors are grateful to the Myelodysplastic Syndromes Foundation for supporting and coordinating this international project.

Conflict of Interest statement. None declared.

REFERENCES

1
Philip
P.
Krogh Jensen
M.
Killmann
S.A.
Drivsholm
A.
Hansen
N.E.
Chromosomal banding patterns in 88 cases of acute nonlymphocytic leukemia
Leuk. Res.
 , 
1978
, vol. 
2
 (pg. 
201
-
212
)
2
Dewald
G.W.
Pierre
R.V.
Phyliky
R.L.
Three patients with structurally abnormal X chromosomes, each with Xq13 breakpoints and a history of idiopathic acquired sideroblastic anemia
Blood
 , 
1982
, vol. 
59
 (pg. 
100
-
105
)
3
Petit
P.
Fryns
J.P.
Masure
R.
Van Den Berghe
H.
Isodicentric (X)(q13): a new characteristic chromosomal anomaly in myeloproliferative syndrome?
Cancer Genet. Cytogenet.
 , 
1982
, vol. 
7
 (pg. 
339
-
341
)
4
Morgan
R.J.
Milligan
D.W.
Williams
J.
Isodicentric X chromosome in a patient with myelodysplastic syndrome
Cancer Genet. Cytogenet.
 , 
1987
, vol. 
27
 (pg. 
215
-
218
)
5
Mackinnon
W.B.
Michael
P.M.
Webber
L.M.
Garson
O.M.
Isodicentric X chromosomes involving the Xq13 breakpoint in myelodysplasia and acute nonlymphocytic leukemia
Cancer Genet. Cytogenet.
 , 
1988
, vol. 
30
 (pg. 
43
-
52
)
6
Temperani
P.
Zucchini
P.
Emilia
G.
Sacchi
S.
Selleri
L.
Torelli
U.
Isodicentric X chromosome in myeloproliferative disorders
Acta Haematol.
 , 
1989
, vol. 
81
 (pg. 
152
-
154
)
7
Chen
Z.
Berger
C.S.
Morgan
R.
Roth
D.
Stone
J.F.
Sandberg
A.A.
Cytogenetic and FISH studies of abnormal X chromosomes in a patient with ANLL
Cancer Genet. Cytogenet.
 , 
1992
, vol. 
62
 (pg. 
130
-
133
)
8
Dierlamm
J.
Michaux
L.
Criel
A.
Wlodarska
I.
Zeller
W.
Louwagie
A.
Michaux
J.-L.
Mecucci
C.
Van den Berghe
H.
Isodicentric (X)(q13) in haematological malignancies: presentation of five new cases, application of fluorescence in situ hybridization (FISH) and review of the literature
Br. J. Haematol.
 , 
1995
, vol. 
91
 (pg. 
885
-
891
)
9
Oscier
D.G.
Atypical chronic myeloid leukaemia, a distinct clinical entity related to the myelodysplastic syndrome?
Br. J. Haematol.
 , 
1996
, vol. 
92
 (pg. 
582
-
586
)
10
Rigolin
G.M.
Cuneo
A.
Roberti
M.G.
Bardi
A.
Castoldi
G.
Myelodysplastic syndromes with monocytic component: hematologic and cytogenetic characterization
Haematologica
 , 
1997
, vol. 
82
 (pg. 
25
-
30
)
11
Martínez-Ramírez
A.
Urioste
M.
Alvarez
S.
Vizmanos
J.L.
Calasanz
M.J.
Cigudosa
J.C.
Benítez
J.
Cytogenetic profile of myelodysplastic syndromes with complex karyotypes: an analysis using spectral karyotyping
Cancer Genet. Cytogenet.
 , 
2004
, vol. 
153
 (pg. 
39
-
47
)
12
Starczynowski
D.T.
Vercauteren
S.
Telenius
A.
Sung
S.
Tohyama
K.
Brooks-Wilson
A.
Spinelli
J.J.
Eaves
C.J.
Eaves
A.C.
Horsman
D.E.
, et al.  . 
High-resolution whole genome tiling path array CGH analysis of CD34+ cells from patients with low-risk myelodysplastic syndromes reveals cryptic copy number alterations and predicts overall and leukemia-free survival
Blood
 , 
2008
, vol. 
112
 (pg. 
3412
-
3424
)
13
McDonell
N.
Ramser
J.
Francis
F.
Vinet
M.C.
Rider
S.
Sudbrak
R.
Riesselman
L.
Yaspo
M.L.
Reinhardt
R.
Monaco
A.P.
, et al.  . 
Characterization of a highly complex region in Xq13 and mapping of three isodicentric breakpoints associated with preleukemia
Genomics
 , 
2000
, vol. 
64
 (pg. 
221
-
229
)
14
Rack
K.A.
Chelly
J.
Gibbons
R.J.
Rider
S.
Benjamin
D.
Lafrenière
R.G.
Oscier
D.
Hendriks
R.W.
Craig
I.W.
Willard
H.F.
, et al.  . 
Absence of the XIST gene from late-replicating isodicentric X chromosomes in leukaemia
Hum. Mol. Genet.
 , 
1994
, vol. 
3
 (pg. 
1053
-
1059
)
15
Sessarego
M.
Bianchi Scarrà
G.
Giuntini
P.
Ajmar
F.
On the Xq13 breakpoint: clinical and cytogenetic observations in a patient with acute myelogenous leukemia
Acta Haematol.
 , 
1983
, vol. 
70
 (pg. 
134
-
136
)
16
Dewald
G.W.
Brecher
M.
Travis
L.B.
Stupca
P.J.
Twenty-six patients with hematologic disorders and X chromosome abnormalities. Frequent idic(X)(q13) chromosomes and Xq13 anomalies associated with pathologic ringed sideroblasts
Cancer Genet. Cytogenet.
 , 
1989
, vol. 
42
 (pg. 
173
-
185
)
17
Barbouti
A.
Stankiewicz
P.
Nusbaum
C.
Cuomo
C.
Cook
A.
Höglund
M.
Johansson
B.
Hagemeijer
A.
Park
S.-S.
Mitelman
F.
, et al.  . 
The breakpoint region of the most common isochromosome, i(17q), in human neoplasia is characterized by a complex genomic architecture with large, palindromic, low-copy repeats
Am. J. Hum. Genet.
 , 
2004
, vol. 
74
 (pg. 
1
-
10
)
18
McCabe
M.G.
Ichimura
K.
Pearson
D.M.
Liu
L.
Clifford
S.C.
Ellison
D.W.
Collins
V.P.
Novel mechanisms of gene disruption at the medulloblastoma isodicentric 17p11 breakpoint
Genes Chromosomes Cancer
 , 
2009
, vol. 
48
 (pg. 
121
-
131
)
19
Mendrzyk
F.
Korshunov
A.
Toedt
G.
Schwarz
F.
Korn
B.
Joos
S.
Hochhaus
A.
Schoch
C.
Lichter
P.
Radlwimmer
B.
Isochromosome breakpoints on 17p in medulloblastoma are flanked by different classes of DNA sequence repeats
Genes Chromosomes Cancer
 , 
2006
, vol. 
45
 (pg. 
401
-
410
)
20
Allikmets
R.
Raskind
W.H.
Hutchinson
A.
Schueck
N.D.
Dean
M.
Koeller
D.M.
Mutation of a putative mitochondrial iron transporter gene (ABC7) in X-linked sideroblastic anemia and ataxia (XLSA/A)
Hum. Mol. Genet.
 , 
1999
, vol. 
8
 (pg. 
743
-
749
)
21
Boultwood
J.
Pellagatti
A.
Nikpour
M.
Pushkaran
B.
Fidler
C.
Cattan
H.
Littlewood
T.J.
Malcovati
L.
Della Porta
M.G.
Jädersten
M.
, et al.  . 
The role of the iron transporter ABCB7 in refractory anemia with ring sideroblasts
PLoS One
 , 
2008
, vol. 
3
 pg. 
e1970
 
22
Payer
B.
Lee
J.T.
X chromosome dosage compensation: how mammals keep the balance
Annu. Rev. Genet.
 , 
2008
, vol. 
42
 (pg. 
733
-
772
)
23
Gondek
L.P.
Tiu
R.
O'Keefe
C.L.
Sekeres
M.A.
Theil
K.S.
Maciejewski
J.P.
Chromosomal lesions and uniparental disomy detected by SNP arrays in MDS, MDS/MPD, and MDS-derived AML
Blood
 , 
2008
, vol. 
111
 (pg. 
1534
-
1542
)
24
Gupta
M.
Raghavan
M.
Gale
R.E.
Chelala
C.
Allen
C.
Molloy
G.
Chaplin
T.
Linch
D.C.
Cazier
J.B.
Young
B.D.
Novel regions of acquired uniparental disomy discovered in acute myeloid leukemia
Genes Chromosomes Cancer
 , 
2008
, vol. 
47
 (pg. 
729
-
739
)
25
Tefferi
A.
Pardanani
A.
Lim
K.-H.
Abdel-Wahab
O.
Lasho
T.L.
Patel
J.
Gangat
N.
Finke
C.M.
Schwager
S.
Mullally
A.
, et al.  . 
TET2 mutations and their clinical correlates in polycythemia vera, essential thrombocythemia and myelofibrosis
Leukemia
 , 
2009
, vol. 
23
 (pg. 
905
-
911
)
26
Delhommeau
F.
Dupont
S.
Della Valle
V.
James
C.
Trannoy
S.
Massé
A.
Kosmider
O.
Le Couedic
J.-P.
Robert
F.
Alberdi
A.
, et al.  . 
Mutation in TET2 in myeloid cancers
N. Engl. J. Med.
 , 
2009
, vol. 
360
 (pg. 
2289
-
2301
)
27
Langemeijer
S.M.
Kuiper
R.P.
Berends
M.
Knops
R.
Aslanyan
M.G.
Massop
M.
Stevens-Linders
E.
van Hoogen
P.
van Kessel
A.G.
Raymakers
R.A.P.
, et al.  . 
Acquired mutations in TET2 are common in myelodysplastic syndromes
Nat. Genet.
 , 
2009
, vol. 
41
 (pg. 
838
-
842
)
28
Iafrate
A.J.
Feuk
L.
Rivera
M.N.
Listewnik
M.L.
Donahoe
P.K.
Qi
Y.
Scherer
S.W.
Lee
C.
Detection of large-scale variation in the human genome
Nat. Genet.
 , 
2004
, vol. 
36
 (pg. 
949
-
951
)
29
Allen
R.C.
Zoghbi
H.Y.
Moseley
A.B.
Rosenblatt
H.M.
Belmont
J.W.
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
, vol. 
51
 (pg. 
1229
-
1239
)
30
Baxter
E.J.
Scott
L.M.
Campbell
P.J.
East
C.
Fourouclas
N.
Swanton
S.
Vassiliou
G.S.
Bench
A.J.
Boyd
E.M.
Curtin
N.
, et al.  . 
Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders
Lancet
 , 
2005
, vol. 
365
 (pg. 
1054
-
1061
)

Author notes

Current address: Signature Genomic Laboratories, 2820 North Astor Street, Spokane, WA 99207, USA.