Elaine M. Sloand, Building 10, Room 7C103, NIH, 9000 Rockville Pike, Bethesda, MD 20892-1652, USA. E-mail: sloande@NIH.gov
Summary. Some patients with paroxysmal nocturnal haemoglobinuria (PNH) have bone marrow findings characteristic of myelodysplastic syndrome. We studied nine PNH patients to determine whether these karyotypic abnormalities were more likely to occur in glycosylphosphatidylinositol-anchored protein (GPI-AP)-negative cells. Abnormal chromosome patterns were evident only in the GPI-AP-positive populations of the PNH clone in 8 of 9 cases studied. Purified GPI-AP-negative CD34 cells gave rise only to cells of normal karyotype, whereas the progeny of the GPI-AP-positive CD34 cells showed the karyotypic abnormality. These findings suggest that environmental factors, but not genetic instability of the GPI-AP-deficient clone, foster development or survival of haematopoietic cells with chromosomal abnormalities.
Paroxysmal nocturnal haemoglobinuria (PNH) is a clonal bone marrow disorder that results from an acquired somatic X-linked mutation of the phosphatidylinositol glycan, class A gene (PIG-A) in a haematopoietic stem cell. The PIG-A gene product functions early in the biosynthetic pathway of the glycosylphosphatidylinositol (GPI). moiety, which anchors many different types of proteins to the cell membrane (Rosse, 1990). All defective erythrocytes, leucocytes and platelets are derived from a single (sometimes multiple) affected haematopoietic progenitor cell(s) carrying a particular PIG-A mutation. Although PNH frequently accompanies aplastic anaemia, bone marrow findings characteristic of myelodysplastic syndrome (MDS), including karyotypic abnormalities and frank dysplasia, may precede or follow its diagnosis; a significant minority of patients initially diagnosed with a myelodysplastic syndrome may show PNH clonal expansion (Longo et al, 1994a; Dunn et al, 1996). PIG-A mutations are insertions or deletions resulting in frameshift mutations. PNH clones have been hypothesized to be more likely to harbour cytogenetic abnormalities or to have an increased probability of leukaemic transformation, because of either intrinsic genomic instability or their greater clonal expansion relative to normal stem cells. These two hypotheses would also explain the greater expansion of the PNH clone relative to normal clones. Consistent with this greater expansion, telomeric length appears to be shortened in GPI-AP-negative cells and may contribute to genetic instability within the PNH clone (our unpublished data). In the mouse telomerase knock-out model (Artandi & DePinho, 2000) and infrequently in individuals deficient in telomerase (dyskeratosis congenita) (Knight et al, 1998; Vulliamy et al, 2001), shortening of telomeres is associated with chromosomal translocations and aneuploidy (Vulliamy et al, 2001). However, experiments with cultured lymphocytes have also suggested that GPI-AP-negative cells might actually be more resistant to mutagenesis than GPI-AP-positive ones, as measured by the hypoxanthine phosphoribosyltransferase assay (Horikawa et al, 2002).
In this study, we sought to determine whether cytogenetic abnormalities occurred more frequently in the GPI-AP-negative clone because of its expansion, relative to GPI-AP-positive clones.
Patients and materials
Patient selection. Blood samples from three healthy volunteers and 9 patients with PNH were obtained after receiving informed consent according to protocols approved by the institutional review board of the National Heart, Lung, and Blood Institute. The diagnosis of PNH was based on clinical signs of intravascular haemolysis and flow cytometric evidence of deficient granulocyte expression of CD16 and CD66. All PNH patients were transfusion dependent, and all had bone marrow failure.
Fluorescent in situ hybridization (FISH). At d 0 and 14, cells were treated with hypotonic buffer consisting of KCl/Hepes, EGTA and NaOH to expose the nucleus at interphase, and the cells were fixed to slides using methanol:acetic acid (3:1). FISH was performed with probes for chromosome 5q, 7 and 8 (Vysis, Downers Grove, IL, USA) as described previously (Sloand et al, 2002a).
Culture conditions. Flow cytometric sorting of GPI-AP-positive and -negative CD34 cells was based on staining by fluorescein isothiocyanate (FITC)-conjugated CD34 monoclonal antibody (mAb) and phycoerythrin (PE)-conjugated CD59 mAb. Numbers of haematopoietic colony-forming cells were measured in methylcellulose colony cultures under standard conditions. Freshly isolated bone marrow mononuclear cells were plated in methylcellulose (Stem Cell Technologies, Vancouver, Canada) in the presence of 50 ng/ml interleukin-3 (IL-3; Genzyme, Boston, MA, USA), 20 ng/ml granulocyte–macrophage colony-stimulating factor (GM-CSF; Boehringer, Indianapolis, IN, USA), 50/ml stem cell factor (Amgen, Thousand Oaks, CA, USA) and 2 U/ml erythropoietin (EPO; Amgen). CD34+ cells were plated at a density of 1 × 105 in 1 ml of medium and 2 ml of methylcellulose in 35 mm dishes for 2 weeks, after which they were subjected to FISH.
Localization of cytogenetic abnormalities to GPI-AP-positive and -negative peripheral blood cells
We studied nine patients with PNH who exhibited karyotypic abnormalities on routine cytogenetic testing. There were four cases of de novo myelodysplastic syndrome (MDS) without a history of aplastic anaemia who developed PNH during the course of their disease, while five had aplastic anaemia/PNH before development of the cytogenetic abnormality (Table I). All had bone marrow failure with hypocellular marrows. We first sorted GPI-AP-positive and -negative cells by flow cytometry based on CD59 expression and then examined for chromosomal abnormalities by FISH on the separated cell fractions. In eight out of nine patients, the karyotypic abnormality was evident only in the GPI-AP-positive population (P < 0·04) (Table I, Fig 1A).
Table I. Association of GPI-AP and cytogenetic abnormalities of CD3-depleted PB mononuclear cells.
Samples of PB from patients with PNH were obtained, red cells lysed and prepared for flow cytometry. Cells were sorted into GPI-AP-positive and -negative fractions based on CD59 expression, and FISH was performed on fractions. In all but one sample, the cytogenetic abnormality was present only in the GPI-positive fraction.
47, XX, 5q–, +8
2% 5q–, 0% trisomy 8
10% 5q–, 30% trisomy 8
47, XX, +8
0% trisomy 8
45% trisomy 8
47, XY, +8
0% trisomy 8
12% trisomy 8
45, XY, −7
18% 7 del
5% 7 del
3% Y del
17% y del
46, XX, 13q–
2% 13 q–
46, XX, +8
3% trisomy 8
15% trisomy 8
45, XY, −11
2% 11 del
32% 11 del
45, XX, −11
1% 11 del
20% 11 del
In the patient whose GPI-AP-negative cells showed monosomy 7 (patient 4), the GPI-AP-positive fraction also showed small numbers of cells scored as monosomy 7 (normal mean = 1 ± 2 cells). In the patients studied (one monosomy 7, one trisomy 8; patients 2 and 4), FISH revealed cytogenetic abnormalities in the GPI-AP-positive CD34 cells. We obtained similar results for two patients (patients 8 and 9) looking at GPI-AP-positive and -negative bone marrow mononuclear cells (data not shown). The cytogenetic abnormality was present in the GPI-AP-positive CD34+ cells and not in those that were GPI-AP-negative.
Chromosome analysis of GPI-AP-positive and -negative cells in tissue culture
In four patients (patients 3, 4, 8 and 9), CD34 cells were sorted based on GPI-AP expression and plated in a methylcellulose short-term proliferation assay, and the resulting haematopoietic colonies were individually subjected to FISH analysis. The PNH phenotype cells gave rise only to progeny of normal karyotype. In contrast, some of the colonies derived from the GPI-AP-positive cells showed the expected karyotypic abnormality. GPI-AP-negative CD34 cells gave rise to significantly greater numbers of erythroid burst-forming units and granulocyte–macrophage colony-forming units (CFU-GM) than did GPI-AP positive CD34 cells (Fig 1B).
Our clinical data suggest that PNH clones are not particularly susceptible to the acquisition of cytogenetic abnormalities. Patients with clonal expansion of GPI-AP-deficient cells with an initial diagnosis of myelodysplasia and those in whom chromosomal abnormalities developed after a history of aplastic anaemia were studied. Only one out of 10 patients showed karyotypic abnormalities within the GPI-AP-negative population. Cytogenetic abnormalities were also present in CD34 cells of the two patients assessed; in both, GPI-AP-deficient CD34 cells gave rise to progeny with normal karyotype, and GPI-AP-positive CD34+ cells gave rise to progeny of which a proportion showed the clinical karyotypical abnormalities. In the one patient whose GPI-AP-deficient cells exhibited monosomy 7, the small percentage of GPI-AP-positive cells with abnormal karyotype may have resulted from contamination during sorting or a PIG-A mutation in a karyotypically abnormal but more committed progenitor cell (with subsequent greater expansion of the PNH clone) or, the chromosome abnormality could have occurred first and the PIG-A mutation later. As we have reported previously for aplastic anaemia/PNH (Chen et al, 2002), GPI-AP-negative CD34 cells produced more colonies than did GPI-AP-positive ones, and we cannot exclude the possibility that the GPI-AP-negative cells gave rise to fewer colonies because of negative in vitro growth characteristics as a result of chromosomal changes. At least two mechanisms could account for karyotypical abnormalities in patients with bone marrow failure. Cells with karyotypical abnormalities or the PIG-A mutation may have a growth advantage over normal cells in bone marrow failure, or they might be intrinsically less likely to undergo apoptosis than are normal cells (Brodsky et al, 1997). Alternatively, clonal abnormalities may become apparent in immune-mediated marrow failure because the genetic changes, mutations or aneuploidy confer some advantage in confronting T-cell attack. Previous evidence from our and other laboratories implies a growth advantage for PNH clones in aplastic anaemia in vivo (Chen, R. et al, 2000; Chen, G. et al, 2002) and in MDS. We reported recently that trisomy 8 cells have a proliferative advantage over normal cells from the same subject (in the presence of Fas antagonist), whereas monosomy 7 cells are less likely to express apoptotic markers (Sloand et al, 2002b). In bone marrow failure states, PNH cells may have a survival advantage in the setting of autoimmunity (Naggakura et al, 2002). Should GPI-AP-positive cells be more susceptible to undergo immune attack (Dunn et al, 1997), mutations chromosomal abnormalities that confer either a proliferative advantage or resistance to apoptosis may be positively selected for in the GPI-AP-positive population and not in the GPI-AP-negative population.
In summary, GPI-AP-negative cells are not more likely to have cytogenetic abnormalities than GPI-AP-positive cells. Cytogenetic abnormalities most probably result from environmental factors that are present in bone marrow failure states, which also favour the emergence of PNH, rather than genetic instability in the GPI-AP-negative clone.