Dominant contribution of malignant endothelial cells to endotheliopoiesis in chronic myeloid leukemia

Experimental Hematology 2009;37:87–91

Dominant contribution of malignant endothelial cells to endotheliopoiesis in chronic myeloid leukemia Jingyi Wu*, Liang Huang*, Mei Huang, Wenli Liu, Miao Zheng, Yang Cao, Yanling Liu, Yicheng Zhang, Yunping Lu, Gang Xu, Shixuan Wang, Ding Ma, and Jianfeng Zhou Cancer Biology Research Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Hubei, P.R. China (Received 13 June 2008; revised 11 August 2008; accepted 27 August 2008)

Objective. Although it has been well-established that hemangioblasts are present in chronic myeloid leukemia (CML) and contribute to both malignant hematopoiesis and endotheliopoiesis, the real contribution of CML-derived endothelial cells to endotheliopoiesis in CML patients has never been evaluated. The current study sought to determine CML-derived endotheliopoiesis in patients with CML. Materials and Methods. Endothelial cells were isolated from the bone marrow or peripheral blood of six newly diagnosed CML patients using an immunomagnetic approach. The resulting endothelial cells were immediately subjected to fluorescence in situ hybridization analysis to determine BCR-ABL-positive endothelial cells. Results. The purity of isolated endothelial cells was 94.47% ± 2.37%. In bone marrow, the BCR-ABL-positive endothelial cells accounted for 70.8% ± 10.7% of total freshly isolated endothelial cells. In peripheral blood, however, the BCR-ABL-positive endothelial cells accounted for only 20.8% ± 9.8% of isolated endothelial cells. Conclusion. The present data demonstrate a dominant contribution of CML-derived endothelial cells to endotheliopoiesis in newly diagnosed CML, and provide the rationale for targeting hemangioblasts and angiogenesis in management of CML. Ó 2009 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc.

Hemangioblasts are bipotent precursor cells that can generate both blood cells and endothelial cells [1]. The presence of hemangioblasts clearly demonstrates the close relationship and interdependence of hemato- and endotheliopoiesis [2]. In the bone marrow, hematopoietic cells are supported by a vascular sinusoidal network [3–5]. Under physiological conditions, the interdependencies of both systems are well-exemplified by the fact that formation of an endothelial network precedes hematopoiesis. Recent mounting evidence has suggested that hemangioblastic precursor cells are present in many types of hematological malignancies, and contribute to maintenance of both malignant hematopoiesis and angiogenesis [6–9]. For example, circulating endothelial cells in multiple myeloma (MM) patients with 13q14 deletion were detected with the same chromosome aberration as the neoplastic plasma cells, suggesting a direct *Drs. Jingyi Wu and Liang Huang contributed to this work equally. Offprint requests to: Ding Ma, M.D., Ph.D. Cancer Biology Research Center, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, 1095 Jiefang Anv. Wuhan, Hubei 430030, P. R. China.; E-mail: [email protected]

contribution of MM-derived circulating endothelial cells to tumor vasculogenesis and possibly to the spreading and progression of MM [8]. In a more recent study, adherent fetal liver kinase-1-positive (Flk1þ) CD31CD34 cells could produce transplantable chronic myeloid leukemia (CML) in severe combined immunodeficient disease (SCID) mice and give rise to both CML blood cells and BCR-ABL-positive endothelial cells at the single cell level, providing direct evidence that cells capable of imitating CML have the potential to produce endothelial cells [10]. Again in CML, endothelial cells generated from bone marrow-derived cells in vitro were found to carry the BCR-ABL fusion gene. More importantly, BCR-ABL-expressing endothelial cells were found in the endothelium of myocardial blood vessels in a CML patient, which suggested that tumor-derived endothelial cells participated in the maintenance of angiogenesis [6]. Despite the recent advancements, little is known about the exact contribution of tumor-derived endothelial cells to endotheliopoiesis in hematological malignancies. Clarification of these questions would help to underscore the role of malignant endotheliopoiesis in the natural course of CML.

0301-472X/09 $–see front matter. Copyright Ó 2009 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc. doi: 10.1016/j.exphem.2008.08.009

J. Wu et al./ Experimental Hematology 2009;37:87–91

88

In the present study, we investigated six patients with newly diagnosed cases of CML for tumor-derived endotheliopoiesis. Endothelial cells were isolated from the bone marrow and peripheral blood using an immunomagnetic approach, and the resulting endothelial cells were immediately subjected to fluorescence in situ hybridization (FISH) analysis to determine BCR-ABL-positive endothelial cells. Our current findings demonstrate that malignant endotheliopoiesis is the dominant endotheliopoiesis occurring in the CML bone marrow.

Materials and methods Patients Bone marrow and peripheral blood samples were collected from six patients with newly diagnosed cases of CML, following a protocol approved by the Institutional Review Board for Human Research of the TongJi hospital, TongJi Medical College, Huazhong University of Science and Technology. Of the six enrolled patients, four were diagnosed in chronic phase (CP) and two in accelerated phase (AP), according to World Health Organization criteria. No chemotherapy or radiotherapy had been performed prior to collection of the samples. The clinical characteristics of the patients are summarized in Table 1. Isolation of endothelial cells Bone marrow aspirates were placed in sterile tubes containing preservative-free heparin (20 U/mL). Every 3 mL bone marrow aspirate was immediately diluted by phosphate-buffered saline (PBS; pH 7.2) at a ratio of 1:1, and layered over Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) for density gradient centrifugation at 400g for 30 minutes. Low-density mononuclear cells at the interface were collected, washed in PBS, and passed through a 50-mm nylon mesh to remove cell clumps. The resultant mononuclear cells were washed twice with PBS (pH 7.2) containing 0.5% bovine serum albumin and 2 mM ethylenediamine tetraacetic acid (buffer A). The mononuclear cells were then resuspended in PBS containing 1% bovine serum albumin at a concentration of 2  107 cells/mL, and incubated with microbeads conjugated to CD105 monoclonal antibody (MACS, Miltenyi Biotech GmbH, Bergisch Gladbach, Germany) at 4 C for 25 minutes. The cells were washed three times with buffer A to remove unconjugated microbeads and loaded on a column in the midiMACS cell separator (Miltenyi Biotech) for separation using a high-gradient

magnetic field. The positive fraction was then loaded onto a new column for a second round of magnetic separation. At the end of the separation, the newly isolated cells were counted and assessed for viability using Trypan Blue dye exclusion. The purity of endothelial cells was determined using a FACS Calibur flow cytometer (Becton Dickinson, San Jose, CA, USA). Isolation and characterization of circulating endothelial cells from 20 mL peripheral blood was performed following the same method as mentioned here. Flow cytometry Purity of isolated endothelial cells was confirmed by doublepositive staining for CD31 [11] and CD105 [12]. Isolated endothelial cells were resuspended in 100 mL RPMI-1640 containing 20% fetal bovine serum at a cell density of 5.0  105 cells/mL. Cells were then incubated with fluorescein isothiocyanate (FITC)-conjugated mouse anti-human CD105 and phycoerythrin (PE)-conjugated mouse anti-human CD31 antibodies (BD Pharmingen, San Jose, CA, USA) for 30 minutes at 4 C, and washed twice with PBS. Cell fluorescence was measured by flow cytometry. Data were analyzed using the CELLQUEST software (Becton Dickinson) and representative of three independent experiments. Immunocytochemistry Immunomagnetically sorted endothelial cells were stained with mouse anti-human monoclonal antibodies against Flk1, von Willebrand factor (VWF) (BD Pharmingen), CD31, CD105, CD34, or CD45 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) [13,14]. Immunophenotypic results were examined under an Olympus fluorescence-equipped microscope (Tokyo, Japan). Data for each sample are given as mean 6 standard deviation (SD) of three slides. FISH To detect the Philadelphia chromosome, FISH was performed with a dual-color, dual-fusion directly labeled LSI BCR-ABL translocation probe (VYSIS, Downers Grove, IL, USA). Cells with two red signals for ABL on chromosome 9 and two green signals for BCR on chromosome 22 were considered normal. Philadelphia chromosome-positive cells showed two yellow fusion signals in addition to one normal ABL (red) and one normal BCR (green) signal. FISH procedures were performed with immunomagnetically sorted endothelial cells according to the manufacturer’s instructions [15]. Visualizing the hybridization was performed using an Olympus fluorescence-equipped microscope with a chargecoupled black-and-white camera device and appropriate hardware

Table 1. Patient characteristics Patient no. Age/gender Phase WBC (103/mL) Hb(g/dL) Plt (103/mL) 1 2 3 4 5 6

54/M 41/M 35/F 20/M 39/M 51/M

AP AP CP CP CP CP

17.2 204.3 61.68 370.61 192.84 169.14

15.2 8.9 4 7.5 7.3 10.9

18 352 56 450 637 686

Karyotype 46,XY,t(9;22)(q34;q11) [6]/47,XY,t(9;22)(q34;q11),þ8 [18] 46,XY,t(9;22)(q34;q11) [26] 46,XX,t(8;9;22)(p23;q34;q11) [20] 46,XY,t(9;22)(q34;q11) [34] 46,XY,t(9;22)(q34;q11) [28] 46,XY,t(9;22)(q34;q11) [20]

FISHþ

a

BCR-ABL (%) 78 87 55 72 69 59

AP 5 accelerated phase; CP 5 chronic phase; F 5 female; FISH 5 fluorescence in situ hybridization; Hb 5 hemoglobin; M 5 male; Plt 5 platelet; WBC 5 white blood cells. a FISH data were derived from bone marrow.

J. Wu et al./ Experimental Hematology 2009;37:87–91

89

Figure 1. Determination of purity of endothelial cells by flow cytometry. Endothelial cells were immunomagnetically isolated from the bone marrow or peripheral blood from chronic myeloid leukemia (CML) patients and subjected to purity analysis by flow cytometry. A shows the background staining of isolated endothelial cells incubated with phycoerythrin-mouse IgG1 isotype control and fluorescein isothiocyanate (FITC)-mouse IgG1 isotype control. B depicts a representative sample assessing endothelial cell purity.

and software. Two-hundred cells were counted for each slide. Data for each sample were given as the mean (6standard deviation) of three slides.

Results To validate the immunomagnetic approach, an immunological phenotyping of the endothelial cells freshly isolated from six CML patients was performed. The purity of yielded endothelial cells, as defined by double-positive staining for CD31 [11] and CD105 [12] (Fig. 1), was 94.47% 6 2.37%. The freshly isolated endothelial cells were further characterized by immunocytochemistry (Table 2, Fig. 2). The percentage of CD45þ cells, an indicator for contamination of white blood cells, was 0.7% 6 0.4%. The

examination using multiple immune markers for endothelial cells (CD31þ, CD34þ, Flk1þ, CD45) [13] consistently demonstrated that the purity of immunomagnetically isolated endothelial cells either of bone marrow or peripheral blood origin was O90% pure. In bone marrow-derived endothelial cells, the mean percentage of VWFþ cells, a marker for mature endothelial cells [14], was 21.3% 6 6.6%, which is in agreement with previous reports that the majority of endothelial cells of bone marrow origin are endothelial progenitor cells [16]. The mean percentage of VWFþ endothelial cells isolated from peripheral blood was 93.3% 6 1.9%, indicating that most of the endothelial cells of peripheral blood origin were mature endothelial cells. Interestingly, there was no obvious increase in peripheral blood in the percentage of circulating endothelial

Table 2. Immunophenotypes and fluorescence in situ hybridization analysis of freshly isolated endothelial cells from bone marrow and peripheral blood of patients with CMLa Patients No. 1 (CML-AP) Bone marrow Peripheral blood No. 2 (CML-AP) Bone marrow Peripheral blood No. 3 (CML-CP) Bone marrow Peripheral blood No. 4 (CML-CP) Bone marrow Peripheral blood No. 5 (CML-CP) Bone marrow Peripheral blood No. 6 (CML-CP) Bone marrow Peripheral blood

CD105þ

CD31þ

CD34þ

Flkþ

VWFþ

CD45þ

BCR-ABLþ

95.2 6 2.8 ND

91.8 6 4.4 ND

96.2 6 2.0 95.7 6 2.3

94.2 6 4.6 ND

14.2 6 3.0 93.7 6 1.6

0.3 6 0.3 ND

81.0 6 9.5 31.0 6 10.1

97.2 6 1.3 97.3 6 0.8

93.5 6 4.0 94.5 6 2.2

95.0 6 3.1 95.5 6 1.3

93.3 6 5.1 96.2 6 1.3

15.7 6 5.5 92.0 6 2.6

0.8 6 1.0 0.3 6 0.6

83.7 6 10.4 35.0 6 6.9

95.0 6 1.5 95.3 6 1.0

95.0 6 3.3 91.8 6 5.3

96.7 6 1.3 94.0 6 4.4

95.3 6 2.6 91.8 6 5.3

22.3 6 8.0 94.5 6 3.1

1.2 6 0.8 0.5 6 0.5

61.7 6 8.5 13.0 6 6.2

92.0 6 5.5 93.5 6 1.8

94.5 6 2.6 95.8 6 2.5

96.3 6 1.5 91.7 6 2.8

96.5 6 0.9 93.2 6 2.6

24.7 6 5.3 90.0 6 4.4

0.5 6 0.5 1.3 6 0.8

57.7 6 9.5 15.7 6 6.8

96.2 6 3.7 94.5 6 4.8

95.5 6 2.2 95.3 6 2.8

95.8 6 1.6 95.8 6 1.6

93.0 6 3.6 96.2 6 0.8

19.0 6 2.3 95.0 6 2.3

0.3 6 0.3 1.2 6 1.0

78.3 6 4.2 18.3 6 9.7

94.8 6 2.9 95.3 6 2.8

92.8 6 4.8 94.5 6 3.3

95.8 6 1.6 93.2 6 3.8

95.2 6 1.9 95.8 6 2.0

32.0 6 4.8 94.7 6 2.4

0.5 6 0.5 0.2 6 0.3

69.7 6 10.6 12.0 6 6.1

AP 5 accelerated phase; CML 5 chronic myeloid leukemia; CP 5 chronic phase; ND 5 not determined; VWF 5 von Willebrand factor. a Values represent percentage of antigen-positive cells out of total isolated cells.

90

J. Wu et al./ Experimental Hematology 2009;37:87–91

Figure 2. Immunophenotypes of freshly isolated endothelial cells from patients with chronic myeloid leukemia (CML). Endothelial cells were immunomagnetically isolated from the bone marrow or peripheral blood of patients with CML. Cells were then stained with mouse antihuman monoclonal antibodies against Flk1, CD34, CD105, CD31, von Willebrand factor (VWF), or CD45, and incubated with fluorescein isothiocyanate (FITC)-conjugated goatanti-mouse secondary antibody. Positive expression was visualized as a green color. (A-F) Representative sections depicting similar data observed in all samples from the six CML patients. Scale bar: 20 mm.

progenitor cells between CML-chronic phase vs CML-AP (Table 2). Next, we determined the proportion of BCR-ABLpositive endothelial cells in freshly isolated endothelial cells from the bone marrow or peripheral blood, using a FISH approach (Fig. 3). In the bone marrow, the BCRABL-positive endothelial cells accounted for 70.8% 6 10.7% of total endothelial cells, indicating that dominant endotheliopoiesis in CML originates from CML hemangioblasts. The level of CML endotheliopoiesis showed an increasing trend when CML was in an accelerated phase. In the peripheral blood, the BCR-ABL-positive endothelial cells accounted for only 20.8% 6 9.8% of total endothelial cells. Notably, an increasing trend of BCR-ABL-positive endothelial cells was observed when CML was in AP (Table 2).

Discussion Although it has been well-established that CML hemangioblasts contribute to both malignant hematopoiesis and endotheliopoiesis, of which malignant endotheliopoiesis forms an essential part of the natural course of CML [6,10,13], the actual contribution of CML-derived endothe-

lial cells to endotheliopoiesis in CML patients has never been evaluated. In this study, we show, for the first time, that dominant endotheliopoiesis occurring in newly diagnosed CML was CML-derived. In addition, the observation that the level of CML endotheliopoiesis showed an increasing trend in two patients with CML-AP suggests a possible positive association of neoplastic endothelial cells and progression of CML, although this needs to be further clarified in future studies. In the current study, the freshly isolated endothelial cells was proven to be highly pure with minor contamination of white blood cells. As we avoided amplifying the sorted endothelial cells in cell culture, the interpretation of our data was rather straightforward. Thus, the current observation might reflect the actual situation of endotheliopoiesis in patients with CML. Our findings strongly support a critical contribution of malignant endotheliopoiesis in the natural course of CML. Recently, Otten et al. [17] has shown that the numbers of outgrowth endothelial cells (OECs) derived ex vivo from CML patients were ninefold higher than those of healthy donors [17]. In the same study, the CML-derived OECs did not harbor the BCR-ABL translocation, and were thus not clonally related to BCR-ABL-positive hematopoietic progenitors. The findings have greatly extended our

J. Wu et al./ Experimental Hematology 2009;37:87–91

91

for development of novel antiangiogenic agents targeting the ‘‘malignant’’ components of endothelium more potently and specifically [18]. Acknowledgments Grant support was received from the National Science Foundation of China (No. 30370657, 30528012, 30600667; 30571950), ‘‘973’’ Program (2002CB513100).

References

Figure 3. Fluorescence in situ hybridization (FISH) analysis of freshly isolated endothelial cells from patients with chronic myeloid leukemia (CML). Immunomagnetically isolated endothelial cells were detected for BCR-ABL translocation with a dual-color, dual-fusion directly labeled LSI BCR-ABL translocation probe. The image depicts a representative result for FISH analysis. The BCR-ABL positive endothelial cell has one red signal for ABL, one green signal for BCR, and two fusion signals (yellow arrowheads). The BCR-ABL-negative endothelial cell has two red signals and two green signals.

understanding of endotheliopoiesis in CML. The phenomenon that OECs and BCR-ABL-positive endothelial cells coexist in a patient with CML is interesting, and implies that the endotheliopoiesis in malignancies is far more complex than previously perceived. Rather than contradicting our present conclusion, Otten et al. [17] have highlighted the presence of a novel hierarchy of endothelial cells that are not clonally related to CML. It should be emphasized that the endothelial cells examined in the two studies were different. In Otten et al.’s system, OECs had been isolated through a culture procedure, while freshly isolated endothelial cells were used in our study. It is possible that BCR-ABL-positive endothelial cells are not maintained or expanded under in vitro conditions. Our findings have important clinical implications. First, based on our current data, CML at diagnosis should be considered to be a disease comprising malignant hemato- and endotheliopoiesis. Second, given the close relationship and interdependence between hemato- and endotheliopoiesis under physiological conditions, targeting angiogenesis or interaction of hemato- and endotheliopoiesis might greatly improve the treatment of CML. Finally, currently available antiangiogeneic agents have been developed based on the understanding of general rather than tumorassociated angiogenesis. Novel antiangiogenic agents targeting tumor-derived endothelial cells rather than genetically normal and stable endothelial cells are needed. Because CML-derived endothelial cells are positive for the BCR-ABL marker, they could be used as an ideal model

1. Choi K, Kennedy M, Kazarov A, et al. A common precursor for hematopoietic and endothelial cells. Development. 1998;125:725–732. 2. Wang L, Li L, Shojaei F, et al. Endothelial and hematopoietic cell fate of human embryonic stem cells originates from primitive endothelium with hemangioblastic properties. Immunity. 2004;21:31–41. 3. Tavassoli M. Structure and function of sinusoidal endothelium of bone marrow. Prog Clin Biol Res. 1981;59B:249–256. 4. Kopp HG, Avecilla ST, Hooper AT, et al. The bone marrow vascular niche: home of HSC differentiation and mobilization. Physiology (Bethesda). 2005;20:349–356. 5. Wilson A, Trumpp A. Bone-marrow haematopoietic-stem-cell niches. Nat Rev Immunol. 2006;6:93–106. 6. Gunsilius E, Duba HC, Petzer AL, et al. Evidence from a leukaemia model for maintenance of vascular endothelium by bone-marrowderived endothelial cells. Lancet. 2000;355:1688–1691. 7. Streubel B, Chott A, Huber D, et al. Lymphoma-specific genetic aberrations in microvascular endothelial cells in B-cell lymphomas. N Engl J Med. 2004;351:250–259. 8. Rigolin GM, Fraulini C, Ciccone M, et al. Neoplastic circulating endothelial cells in multiple myeloma with 13q14 deletion. Blood. 2006;107:2531–2535. 9. Della Porta MG, Malcovati L, Rigolin GM, et al. Immunophenotypic, cytogenetic and functional characterization of circulating endothelial cells in myelodysplastic syndromes. Leukemia. 2008;22:530–537. 10. Fang B, Zheng C, Liao L, et al. Identification of human chronic myelogenous leukemia progenitor cells with hemangioblastic characteristics. Blood. 2005;105:2733–2740. 11. Baldwin HS, Shen HM, Yan HC, et al. Platelet endothelial cell adhesion molecule-1 (PECAM-1/CD31): alternatively spliced, functionally distinct isoforms expressed during mammalian cardiovascular development. Development. 1994;120:2539–2553. 12. Gougos A, Letarte M. Identification of a human endothelial cell antigen with monoclonal antibody 44G4 produced against a pre-B leukemic cell line. J Immunol. 1988;141:1925–1933. 13. Prindull G. Hemangioblasts representing a functional endotheliohematopoietic entity in ontogeny, postnatal life, and CML neovasculogenesis. Stem Cell Rev. 2005;1:277–284. 14. Wagner DD. Cell biology of von Willebrand factor. Annu Rev Cell Biol. 1990;6:217–246. 15. Schlegelberger B, Metzke S, Harder S, et al. Classical and molecular cytogenetics of tumor cells. In: Wegner RD, ed. Diagnostic Cytogenetics. New York, Berlin, Heidelberg: Springer-Verlag; 1999. p. 151–185. 16. Dome B, Dobos J, Tovari J, et al. Circulating bone marrow-derived endothelial progenitor cells: characterization, mobilization, and therapeutic considerations in malignant disease. Cytometry A. 2008;73: 186–193. 17. Otten J, Schultze A, Schafhausen P, et al. Blood outgrowth endothelial cells from chronic myeloid leukemia patients are BCR/ABL1 negative. Br J Haematol. 2008;142:115–118. 18. Kvasnicka HM, Thiele J, Staib P, et al. Reversal of bone marrow angiogenesis in chronic myeloid leukemia following imatinib mesylate (STI571) therapy. Blood. 2004;103:3549–3551.