| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Division of Cellular Signaling, Institute for Advanced Medical Research, Keio University School of Medicine, Tokyo 160-8582, Japan [H. I., M. M., Y. Kawak.]; Department of Clinical Genetics and Human Genetics, VU University Medical Center 1081 BT, Amsterdam, the Netherlands [Q. W., A. B. O., A. W. M. N., J. P. d. W., H. J.]; Department of Pediatrics, Shizuoka Red Cross Hospital, Shizuoka 420-0853, Japan [A. K.]; Division of Molecular Medicine, Oregon Health and Science University, Portland, Oregon 97201 [M. E. H.]; Department of Laboratory Medicine, Keio University School of Medicine, Tokyo 160-8582, Japan [Y. Kawai]; Radiation Biology Center, Kyoto University, Kyoto 606-8501, Japan [M. S. S.]; and Department of Pediatric Oncology, Dana Farber Cancer Institute, and Department of Pediatrics, Children’s Hospital, Harvard Medical School, Boston, Massachusetts 02115 [A. D. D.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Eight complementation groups have been described in FA (2, 3, 4) . The genes corresponding to groups A, C, D2, E, F, and G have been cloned (5 , 6) , whereas the gene defective in groups B and D1 has been identified as BRCA2 (7) . The FA proteins FANCA, FANCC, FANCE, FANCF, and FANCG form a functional multiprotein complex in the nuclear compartment. The nuclear FA protein complex is required for the activation of the FANCD2 protein into a monoubiquitinated isoform, which colocalizes and interacts with BRCA1 in DNA-damage-inducible nuclear foci (8) . With the recent identification of BRCA2 as a FA gene, a picture is emerging of an integrated FA/BRCA nuclear caretaker pathway that protects against the disease features of FA and development of specific malignancies, including AML.
The predisposition of FA patients to malignancies presumably is related to the chromosomal instability feature of the syndrome, but it is unclear why the malignancies mainly involve AML and squamous cell carcinomas. In FA the relative risk of AML, which has been reported to be as high as 15,000x (9)
, contributes to a strongly reduced average life expectancy, which is currently 
20 years (1)
.
There is no firm evidence that the distribution of AML subtypes according to the French-American-British classification would be different from that seen in the general population. The most important difference seems to be the age at diagnosis; 64 years in the general population as opposed to 14.8 years in FA patients (10) . FA patients with AML have a poor prognosis, with a mean age of death of 15.5 years (9) . The reason why treatment of AML in FA patients generally fails is unknown but may relate to specific properties of the leukemic cells and/or to a reduced capacity of the FA patient to tolerate the chemotherapy regimens used. For answering questions related to AML in both FA and non-FA leukemia patients, the availability of stably growing AML cell lines is crucial. A number of cell lines, derived from various French-American-British subclasses of non-FA AML patients, have been described in the literature (11) . Here, we report the establishment and partial characterization of the first AML cell line derived from an FA patient.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Establishment of a Stably Growing AML Cell Line.
A heparinized peripheral blood sample taken during the second relapse was subjected to culture in February 1998. Mononuclear cells were isolated with Lymphoprep (Nycomed) gradient centrifugation. The cells were washed in PBS and then seeded into 24-well tissue culture flasks (Sumiron MS80240) at 106 cells/ml with 2 ml of RPMI 1640 (Life Technologies, Inc., Gaithersburg, MD) supplemented with 10% FCS. Additional supplementation of the medium was with either 10 ng/ml GM-CSF or 10 ng/ml IL-3. Cultures were kept at 37°C under a 5% CO2-humidified atmosphere. The cells were fed once or twice weekly by partial replacement of spent medium with fresh medium supplemented with GM-CSF or IL-3. Without cytokines, the cells stopped growing in a few days and gradually died. However, continuous growth occurred with GM-CSF or IL-3. These cells could be cryopreserved and successfully recultured after thawing. We designated the cell line FA-AML1. When subjected to routine testing for Mycoplasma contamination, FA-AML1 cultures appeared positive for Mycoplasma fermentans. Therefore, two more cell lines were established from frozen samples of the original leukemic blasts (FA-AML1A, cultured with GM-CSF, and FA-AML1C, cultured with IL-3), which were found to be free of Mycoplasma. Cell line SKNO-1, derived from a non-FA AML (M2) patient [kindly provided by Dr. Noboru Fujinami (SRL, Tokyo, Japan)], was used as a control (14)
. SKNO-1 cells were cultured in RPMI 1640 with 10% FCS, supplemented with 10 ng/ml GM-CSF. The cell lines used in this study are summarized in Table 1
.
|
Hemopoietic Growth Factors.
Recombinant human IL-3, IL-6, GM-CSF, granulocyte colony-stimulating factor, EPO, stem cell factor, and TPO were kindly provided by Kirin Brewery Co. (Tokyo, Japan). IL-4 was a gift of Ono Pharmaceutical Co. (Tokyo, Japan). IFN-
was a gift of Shionogi Pharmaceutical Co. (Tokyo, Japan). M-CSF was a gift of Morinaga Milk Co. (Kanagawa, Japan).
Growth Stimulation Assays.
Cell proliferation in short-term culture was estimated by a modified version of the colorimetric WST-1 assay. Cells cultured with growth factors were washed once and were resuspended in RPMI 1640 with 10% FCS for 24 h. Cells (2 x 104) were incubated with various cytokines in 0.1 ml for 4–6 days in 96-well microculture plates. At the completion of culture, the starting number of the cells were prepared as reference cells. Ten µl of 3.3 mg/ml WST-1 (Dojindo Co., Kumamoto, Japan) dissolved in 0.2 mM 1-methoxy-5-methylphenazinium methylsulfate with 20 mM of HEPES (pH 7.4) was added to each culture well. After 2 h of incubation with WST-1 at 37°C, the absorbance (A) was measured using a microplate reader (model Benchmark; Bio-Rad) at a wavelength of 450 nm with reference wavelength of 655 nm. Percentage of growth was calculated as (sample A)/(reference cells A) x 100%.
Cytogenetic Analysis.
Bone marrow cultures and FA-AML1 cells that had gone through 9, 33, or 60 passages, respectively, were subjected to standard cytogenetic analysis of trypsin-Giemsa-stained metaphase preparations, which were made after a 4-h (37°C) Colcemid treatment. Karyotypes were described according to the International System for Human Cytogenetic Nomenclature (15)
. Fluorescence in situ hybridization with a total chromosome DNA probe for chromosome 7 (Oncor, Inc., Illkirch Cedex, France) and with a probe for the Williams Syndrome region on chromosome 7 [LSI Williams Syndrome (Elastin Gene) Region Probe; Vysis Inc., Downers Grove, IL] was performed according to the manufacturers’ instructions.
MMC-induced chromosomal breakage was assessed in FA-AML1A and SKNO-1 cells by inspection of Giemsa-stained metaphase spreads, as described previously (12 , 16) .
MMC-induced Growth Inhibition.
The growth-inhibiting effect of MMC was assessed by growing the cells in the presence of various concentrations of MMC over a period of time that allowed cells without addition of drug to undergo at least three populations doublings (typically 4–8 days), as described previously (17)
. IC50s are defined as the concentration of MMC causing 50% inhibition of growth.
Mutation Analysis.
Mutation screening in FANCA was reported previously (13)
. FANCC, FANCE, FANCF, and FANCG were screened for mutations by sequencing fragments amplified from genomic DNA. Mutation screening in BRCA2 was by denaturing gradient gel electrophoresis, essentially as described previously (18)
; primers were obtained from Ingeny (Leiden, the Netherlands), and mutations were identified by sequencing of fragments found to be aberrant.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
90%. When cultured in RPMI 1640 supplemented with 10% FCS without additional growth factors, the cells ceased to proliferate within 2–3 days. However, with a supplement of 10 ng/ml GM-CSF or IL-3, the cells continued to proliferate for >20 passages and were therefore considered a permanent cell line, termed FA-AML1. After 28 passages, a clone was selected (B2) by limiting dilution of FA-AML1 cells. The clone could grow without feeder cells and clonality was confirmed by cytogenetic analysis (Table 2)
. B2 cells expressed both GM-CSF and IL-3 receptors, as analyzed by fluorescence-activated cell sorting, and after 40 passages in medium with GM-CSF, they proliferated equally well with IL-3 (data not shown). FA-AML1 cells were maintained in suspension culture for >24 months (>60 passages), with a population doubling time between 30 and 48 h and were thus considered the first successfully established permanent AML cell line from a FA patient.
|
3% of the cells, whereas 
-naphthyl butyrate esterase and naphthol AS-D chloroacetate esterase staining were 100% negative. Electron microscopy confirmed that most of the cells had considerable nuclear segmentation with multiple nucleoli. Furthermore, they showed well developed primary granules, mitochondria, and endoplasmic reticulum in the cytoplasm (results not shown). The morphology of the FA-AML1 cells did not seem to be distinct or unique in any particular respect when compared with other leukemic cell lines.
Cytogenetic Studies.
On admission, cytogenetic analysis of bone marrow samples revealed 15 of 17 cells to be aberrant by showing a ring chromosome 7 in addition to several other aberrations (Table 2)
. Ring 7 persisted in all of the samples examined during subsequent follow-up, whereas subsequently, an addition of the long arm of chromosome 18 (add18q23) appeared in most of the clinical samples. These two aberrations in addition to several others were also consistently observed in FA-AML1 cells after 9, 33, and 60 passages in the clonally derived B2 cells (Table 2)
, indicating that the cell line was cytogenetically rather stable and that it closely resembled the primary leukemic blasts. The independently established cell lines FA-AML1A and FA-AML1C appeared to have a grossly similar karyotype that included ring 7 and several of the marker chromosomes, indicating that they had originated from the same leukemic blast population as FA-AML1. Fluorescence in situ hybridization with a whole chromosome probe for chromosome 7 showed that the ring originated from a chromosome 7. With the Williams Syndrome probe set, which hybridizes to bands 7q11 and 7q31, we demonstrated that most of the long arm was present in the ring. The short arm appeared to be missing.
Surface Marker Analysis.
Cell surface markers of FA-AML1 cells were analyzed and compared with primary leukemic cells from the bone marrow (Table 3)
. Both the primary leukemic cells and FA-AML1 cells were positive for CD11b, CD13, CD33, CD34, and CD38, suggesting that the cells were presumably derived from colony-forming unit, granulocyte-macrophage myelomonocytic stem cells. All of the erythroid-, platelet-, and lymphoid-associated antigens were negative. Most of the surface antigens present in the primary cells were also present in the cell line, except for Ia (HLA-DR), which was strongly positive in the primary cells but absent from the cell line. The adhesion molecule CD11b was strongly positive in FA-AML1 cells but not determined in the primary leukemic sample. These results indicated a myeloid origin of the leukemic cells with a fair degree of preservation of surface antigen expression in the established cell line.
|
, whereas in parallel experiments M-CSF, granulocyte colony-stimulating factor, TPO, and TPA failed to support a sustained proliferative response. When tested in a [3H]thymidine incorporation assay over a 24-h time period, both GM-CSF and IL-3 induced a significant incorporation of label, whereas in the same experiment EPO was ineffective (Fig. 1B)
. Very similar responses were noted for the B2 clonal cell line derived from FA-AML1 cells (results not shown). GM-CSF or IL-3 were necessary but not sufficient for supporting the growth in the media used, because without the FCS or human AB serum supplement, the cells failed to proliferate.
|
,25-dihydroxy vitamin D3 (100 nM) during a 7-day culturing period in RPMI 1640 with 10% FCS. None of the agents appeared to have a detectable effect on cell number or cellular morphology, except for TPA, which induced a spindle shape in a minority of the cells. In addition, TPA at 5 nM supported a weak proliferative response, comparable with TPO and M-CSF (Fig. 1A)
.
Absence of the FA Phenotype in FA-AML1 Cells.
To assess whether FA-AML1 cells expressed the FA phenotype, we determined their sensitivity to growth inhibition by MMC. As shown in Table 4
, IC50s for the various sublines of FA-AML1 cells were variable and ranged between 10.1 and 36.5 nM MMC, with an average for all observations of 24.1 nM MMC (n = 14). This value was lower than that for SKNO-1 (non-FA AML) cells but higher than typically observed for EBV-immortalized lymphoblasts from FA patients (Table 4)
, suggesting that FA-AML1 cells were not as sensitive to MMC as might be expected for FA cells. We then tested their sensitivity to chromosomal breakage by MMC. Table 5
shows the clastogenic effect of MMC in the patient’s T lymphocyte cultures, B lymphoblasts, and AML cells. In contrast to the patient’s T lymphocytes and B lymphoblasts, which exhibited a FA-like hypersensitivity to MMC, the AML cells responded as normal. We repeated the assay by comparing FA-AML1 cells with SKNO-1 cells and lymphoblasts from an established FA group D1 patient (HSC62), which confirmed that FA-AML1 cells responded as non-FA (Fig. 2)
. These results indicated that the FA-AML1 cell line did not express an FA phenotype, suggesting that during the development of AML the FA phenotype had reverted to wild type.
|
|
|
. Thus, these inherited mutations were most likely the underlying cause of the disease in the present FA patient.
|
To determine whether the secondary (correcting) mutation had occurred relatively early in the genesis of the leukemia, we tested frozen peripheral blood leukocytes collected at diagnosis, i.e., before the bone marrow transplant, which had a blast count of 45%. As shown in Fig. 3B
, the secondary mutation was not detectable in these cells, although the mutation was readily detectable in samples that contained 50% of FA-AML1 DNA, which harbors this mutation. This result indicated that the secondary mutation had occurred during a relatively late stage in the progression of the disease. The possibility that the secondary mutation had originated in vitro was considered highly unlikely because the secondary mutation was present in three independently established AML cell lines (FA-AML1, FA-AML1A, and FA-AML1C; see Fig. 3C
). Interestingly, FA-AML1A cells appeared to be a mixed population of cells that contained either the original mutation or the reverted allele with the secondary mutation. This is in agreement with the relatively low IC50s observed for this cell line (Table 4)
.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Chromosomal analysis revealed a ring 7 in which the short arm appeared to be missing. This was the sole abnormality in about one-third of the cells analyzed in the first sample, and this primary change persisted in all of the abnormal cells in all samples investigated subsequently. One of the most frequent numerical aberrations myelodysplastic syndrome found in patients with and AML is monosomy 7, which is not specific for M2 or any other subtype. The frequent involvement of chromosome 7 in clonal abnormalities in AML in patients with FA has been summarized in the literature (22) . In most FA cases, the whole of one chromosome 7 is absent, but cases in which the short arm of chromosome 7 is missing have also been described [deletion 7p (23) ; isochromosome 7q (24) ].
Given the strong predisposition of FA patients to AML, the question arises as to whether AML cells occurring in FA are in any way different from those occurring in non-FA leukemia patients. The morphological, immunological, or cytogenetical characteristics of FA-AML1 cells have failed to reveal any distinct features. Remarkably, the hypersensitivity to MMC that is characteristic for FA cells was not expressed in the FA-AML1 cells in terms of both drug-induced growth inhibition and chromosomal breakage rates or cell cycle arrest. Absence of the FA trait (G2 arrest) has been reported in blood samples from 3 of 4 FA patients who presented with overt leukemia (25) . In addition, progressive disappearance of initially observed G2 arrest was observed in a FA patient during development of AML (26) . These observations suggest that loss of the FA trait may be a common feature of AML developing in FA patients.
Loss of the cellular FA phenotype is a well-known phenomenon occurring in primary lymphocytes from patients with somatic mosaicism. Such mosaicism occurs in 20–30% of FA cases and is caused by acquisition of a functional allele at the disease locus because of secondary mutations or mitotic recombination, which apparently renders these cells a proliferative advantage over nonreverted cells (19
, 21)
. We have found here that a similar mechanism, functional correction of a mutated BRCA2 allele by a secondary mutation, accounts for the loss of a cellular FA phenotype in AML cells. Unfortunately, no material was available to determine whether the secondary mutation was present in primary cells. However, the fact that the same secondary mutation was detected in the three independently established AML cell lines strongly suggests that the event had occurred in vivo. Because the secondary mutation was not detected in early samples, taken before the bone marrow transplant, this mutational event may represent a relatively late growth-promoting step in the progression of the disease. In addition to providing the leukemic cells with more stable growth characteristics, reversion at the disease locus may be an important contributing factor in the development of drug resistance, particularly against cross-linking agents, in later stages of the disease.
The phenomenon of genetic reversion in leukemia developing in FA patients may be more widespread because flow cytometry of AML-containing blood samples have repeatedly resulted in false-negative FA diagnoses. However, additional leukemic FA patients, preferably with defects in known FA genes, should be studied to test this hypothesis. Because the reverted AML cells in our patient were only detected after several chemotherapy treatments, treatment may have played a role, either by directly generating the mutation de novo or by positive selection of reverted cells that preexisted at a low level that escaped detection.
Our findings may have important consequences for the treatment of AML in FA patients where the use of cross-linking cytostatic agents such as cyclophosphamide may be considered a risk factor. Because genetic correction of the FA defect in (pre)leukemic cells in FA patients presumably will promote their growth potential and drug resistance, gene therapy trials intended to correct the marrow failure in FA patients may hold a leukemogenic risk in cases where preexisting (pre)leukemic cells may become targets of the therapeutic vector. Finally, our results highlight a potentially general mechanism of carcinogenesis in which an initial defect in a caretaker gene is spontaneously corrected after the necessary genomic alterations have been accumulated.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by the Netherlands Organization for Scientific Research Grant B 92-209. 
2 Present address: Department of Pathology, Sapporo Medical University School of Medicine, S-1, W-17, Chuo-ku, Sapporo 060-8556, Japan.
3 To whom requests for reprints should be addressed, at Department of Clinical Genetics and Human Genetics, VU University Medical Center, Van der Boechorststraat 7, 1081 BT Amsterdam, the Netherlands. E-mail: h.joenje.humgen{at}med.vu.nl
4 The abbreviations used are: FA, Fanconi anemia; AML, acute myeloid leukemia; MMC, mitomycin C; IL, interleukin; GM-CSF, granulocyte macrophage colony-stimulating factor; FA-AML1, FA-derived AML 1; EPO, erythropoietin; TPO, thrombopoietin; M-CSF, macrophage colony-stimulating factor; WST-1, 2-(4-iodophenyl)-3-(4-nitro-phenyl)-5-(2,4-disulfophenyl)-2H tetrazolium, sodium salt; TPA, 12-O-tetradecanoylphorbol-13-acetate.
Received 11/ 6/02. Accepted 3/19/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Gallmeier and S. E. Kern Targeting Fanconi Anemia/BRCA2 Pathway Defects in Cancer: The Significance of Preclinical Pharmacogenomic Models Clin. Cancer Res., January 1, 2007; 13(1): 4 - 10. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Soulier, T. Leblanc, J. Larghero, H. Dastot, A. Shimamura, P. Guardiola, H. Esperou, C. Ferry, C. Jubert, J.-P. Feugeas, et al. Detection of somatic mosaicism and classification of Fanconi anemia patients by analysis of the FA/BRCA pathway Blood, February 1, 2005; 105(3): 1329 - 1336. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Wagner, J. Tolar, O. Levran, T. Scholl, A. Deffenbaugh, J. Satagopan, L. Ben-Porat, K. Mah, S. D. Batish, D. I. Kutler, et al. Germline mutations in BRCA2: shared genetic susceptibility to breast cancer, early onset leukemia, and Fanconi anemia Blood, April 15, 2004; 103(8): 3226 - 3229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Hirsch, A. Shimamura, L. Moreau, S. Baldinger, M. Hag-alshiekh, B. Bostrom, S. Sencer, and A. D. D'Andrea Association of biallelic BRCA2/FANCD1 mutations with spontaneous chromosomal instability and solid tumors of childhood Blood, April 1, 2004; 103(7): 2554 - 2559. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Warren, C. J. Lord, J. Masabanda, D. Griffin, and A. Ashworth Phenotypic effects of heterozygosity for a BRCA2 mutation Hum. Mol. Genet., October 16, 2003; 12(20): 2645 - 2656. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. van der Heijden, R. H. Hruban, and S. E. Kern Reply Cancer Res., October 15, 2003; 63(20): 6999 - 7001. [Full Text] [PDF]
|
|
|
|
 |
|
 A. D. D'Andrea The Fanconi road to cancer Genes & Dev., August 15, 2003; 17(16): 1933 - 1936. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
