Cell Division Rates of Primary Human Precursor B Cells in Culture Reflect In Vivo Rates
http://www.100md.com
《干细胞学杂志》
a Pathology, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA;
b Pathology and Laboratory Medicine, University of Pennsylvania, Pennsylvania, USA;
c Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
Key Words. Cell division ? Precursor B cell ? Lymphoblastic leukemia ? Stroma-based culture
Correspondence: John Kim Choi, M.D., Ph.D., Children’s Hospital of Philadelphia, 802F ARC, 3516 Civic Center Boulevard, Philadelphia, Pennsylvania 19104, USA. Telephone: 215-590-7194; Fax: 215-573-0342; e-mail: jkchoi@mail.med.upenn.edu
ABSTRACT
Primary normal and leukemic human precursor B cells can be cultured on bone marrow stromal cell layer for at least 3–4 weeks . Studies using this model have provided important insights into the cell pathways that promote the expansion of human precursor B cells. For example, expansion of normal or leukemic human B cells does not require interleukin-7 (IL-7) , a cytokine necessary for the expansion of murine precursor B cells. The rate of expansion is dictated, in part, by a balance of apoptosis and cell division. Apoptosis in both normal and leukemic precursor B cells is inhibited by direct contact with stromal cells and is mediated by vascular cellular adhesion molecule-1, 4?1 integrins, and multiple antiapoptotic proteins . Although the rate of apoptosis is similar among different samples of normal human precursor B cells, primary leukemic precursor B cells have variable rates of apoptosis; decreased apoptotic rate in culture is an independent predictor of poor clinical outcome .
Cell division is also an important contributor to precursor B cell expansion and may occur at a high rate even without expansion if there is offsetting apoptosis and differentiation. Studies of freshly isolated bone marrow cells demonstrate that the rates of cell division are similar among normal human precursor B cells but variable among leukemic precursor B cells. Furthermore, low rates of cell division in leukemic cells may correlate with poor response to chemotherapy and poor clinical outcome . As with apoptosis, stroma-based cultures provide a potentially powerful model for studying cell pathways that regulate cell division of primary normal and leukemic human precursor B cells. However, cell division has not been well characterized in the stroma-based culture model, and it is unknown how well precursor B cells divide in culture. In this study, we followed cell division using the fluorescent dye carboxyfluorescein diacetate, succinimyl ester (CFSE). This approach has not been previously used to study human lymphocytes but has been used to quantify cell division in murine lymphocytes . We report that different patient samples of normal precursor B cells have a relatively constant rate of cell division that is similar to reported in vivo rates. Cultured primary human leukemic cells have more heterogeneous rates of cell division that correlate with their in vivo proliferation index. These findings indicate that stroma-based cultures and CFSE can be used to study cell division of cultured primary human precursor B cells.
MATERIALS AND METHODS
Isolation and Characterization of Primary Normal Human CD19+ B Cells
Progenitor B cells were isolated by Ficoll gradient followed by using antibodies against CD19. Flow cytometry analysis of the purified cells showed >95% of the cells were positive for the B cell–specific antigen CD79a; of these, 70%–100% were positive for the precursor B cell antigen CD10 (Fig. 1). The CD19+, CD10+ cells were negative for the more mature B cell antigen, surface immunoglobulin heavy chain (IgM), and surface immunoglobulin light chains. The remaining 0%–30% of the CD19-selected cells was more mature CD79a+, CD10–, IgM+ B cells. More than 98% of the CD19+ cells were viable by 7-AAD exclusion (data not shown).
Figure 1. Characterization of isolated CD19+ B cells. Mononuclear bone marrow cells, before and after immunomagnetic selection with CD19 antibodies, were stained with phycoerythrin-conjugated anti-CD79a and allophycocyanin-conjugated anti-CD10 and analyzed by flow cytometry.
The CD19+ B cells were cocultured on stromal cells in X-vivo 10 media supplemented with 5% FBS and the growth factors IL-7, SCF, IGF-1, and FLT-3, as previously described . The AFT024 murine stromal cell line was replaced by primary pediatric bone marrow stromal cells, because in our hands the irradiated AFT024 cells continued to proliferate and overgrew the culture after 2 weeks (data not shown). At 0, 1, 2, and 3 weeks, the cells were trypsinized, counted, and analyzed by flow cytometry for CD19 and CD10 expression. Studies using a leukemic precursor B cell line, Nalm-6, indicated that trypsinization up to 15 minutes had no effect on the detection of CD19 and CD10 (data not shown). The expansions of CD19+, CD10+ precursor B cells (n = 6) were variable among different samples, with the cell numbers being zero to four times the initial number of plated cells at the end of 3 weeks. Most samples increased in number during the first 2 weeks and then decreased in number during the third week (Fig. 2). The variable expansions among samples and the decreased numbers after 2 weeks in culture are similar to the findings reported by others .
Figure 2. Expansion curves of CD19+, CD10+ precursor B cells from six patients. Isolated CD19+ B cells were cultured on bone marrow stroma. After 1, 2, and 3 weeks of culture, precursor B cells were enumerated by multiplying the total number of cells by the percentage of CD19+, CD10+ cells that was determined by flow cytometry analysis. The numbers were normalized to the numbers at initial plating.
Expansion of Cultured B Cells Is Determined by a Balance of Cell Division and Apoptosis
Two models can explain the variable, and in some cases lack of, expansion of precursor B cells among different samples with prolonged culture. In the first model, human B cells in some samples divide very slowly or not at all during prolonged culture. This predicts that cultures with no increase or net decrease in B cells have no cell division. In the other model, B cells are dividing but do not expand in number because of concomitant apoptosis. This predicts that all cultures contain dividing B cells. To determine which model is more accurate, we analyzed our cultures for active DNA synthesis by BrdU incorporation. Three-week-old cultures that had a net loss in B cells were incubated with BrdU, and the incorporated BrdU was detected using fluorescein-conjugated monoclonal antibodies to BrdU. UV microscopy showed many positive nuclei (Fig. 3), indicating that even in cultures without obvious B cell expansion, there was active cell division. The dividing cells were likely B cells, because the stromal cells were irradiated and thus unlikely to divide. In support, 3-week-old cultures of irradiated stromal cells without B cells demonstrated no BrdU incorporation (data not shown).
Figure 3. DNA synthesis and apoptosis in culture. Three-week-old cultures were incubated without (top row) and with (middle row) BrdU for 16 hours, stained with anti-BrdU antibodies, and examined under phase and UV microscopy (magnification x200). Three-week-old cultures were analyzed by terminal deoxynucleotidyl transferase–mediated rhodamine-dUTP nick-end labeling (TUNEL) and examined under phase and UV microscopy (magnification x 200).
Because there was a net decrease in B-cell number in some cultures despite DNA synthesis, apoptosis was likely present in our culture. To confirm this, 3-week-old cultures were examined for DNA fragmentation by TdT-mediated rhodamine-dUTP nick-end labeling (TUNEL). UV microscopy showed many positive nuclei (Fig. 3), confirming the presence of apoptosis. Three-week-old cultures of irradiated stromal cells without B cells showed virtually no TUNEL activity, suggesting that the dying cells were of B-cell origin (data not shown). These studies indicate that B-cell expansion in culture is dictated by a balance of cell division and apoptosis.
Rates of Cell Division in Different Samples of Cultured Normal Human B Cells Are Similar
Next, we asked if the variable expansions of precursor B cells could be explained by different rates of cell division among different samples. To quantify the rate of cell division, isolated human CD19+ B cells were labeled with CFSE, cultured, and followed for decreasing CFSE signal. In this approach, CFSE is covalently bound to intracellular macromolecules and eventually produces a stable fluorescent signal for each cell. With each division, the resulting daughter cells have half of the CFSE signal . At 0, 1, 2, and 3 weeks of culture, the cells were trypsinized, stained with antibodies against CD19 and CD10, and analyzed by three-color flow cytometry (Fig. 4A). Our CFSE analysis of human precursor B cells did not produce individual fluorescent peaks as seen with CFSE analysis of murine peripheral T lymphocytes . The lack of individual peaks most likely represented the heterogeneous sizes of the precursor B-cell population leading to overlapping peaks. Backgating on CD19+, CD10+ B cells confirmed a wide size distribution as determined by forward scatter; in contrast, the more mature CD19+, CD10– B cells showed more uniform size (data not shown). Nevertheless, this assay permitted a good estimation of the number of cell divisions.
Figure 4. CFSE analysis of normal human precursor B cells. (A): Isolated CD19+ B cells were labeled with CFSE and cultured on bone marrow stroma. After 4, 7, 14, and 21 days of culture, cells were stained with phycoerythrin-conjugated anti-CD19 and allophycocyanin-conjugated anti-CD10. Cells in the blast and lymphocyte region by forward and side scatter were subgated into CD19+, CD10+ and CD19+, CD10– B cells and then analyzed for CFSE expression. Figure is representative of six samples. (B): Same sample as in (A), except the growth factors were removed. Figure is representative of three samples. Abbreviations: CFSE, carboxyfluorescein diacetate, succinimyl ester; MFI, mean fluorescent intensity
Under our culture conditions, the more mature CD19+, CD10– B cells had a stable CFSE signal that decreased by only 25% over the 3 weeks of culture, suggesting little if any cell division within this population. The constant CFSE signal in this population confirmed the stable nature of the CFSE signal during prolonged culture. The constant CFSE signal also argued against significant differentiation of the dividing precursor cells, because this also results in CD19+, CD10– B cells with decreased CFSE signal. Hence, most of the CD19+, CD10– cells after 3 weeks in culture represented the initial plated cells.
Unlike the more mature CD10– B cells, cultured CD19+, CD10+ precursor B cells showed continuous decrease in CFSE signal throughout the 3 weeks of culture, indicating continuous cell division throughout the culture (Fig. 4). The rate of decreasing CFSE signal was examined for six different samples at weekly intervals (Fig. 5). The CFSE signals were normalized to the stable signal of the CD19+, CD10– B cells. This normalization compensated for the inherent variations in CFSE labeling between samples and flow cytometry settings between weeks. The samples had similar rates of cell division despite the variability in the expansion. During week 1 of culture, the average CFSE signal decreased by 3.14-fold, corresponding to approximately 1.65 divisions. During week 2 of culture, the CFSE signal decreased further by 3.88-fold, corresponding to approximately 1.95 divisions. During week 3 of culture, the CFSE signal decreased further by 3.90-fold, corresponding to approximately 1.89 divisions. Over the 3 weeks of culture, the CFSE signal decreased by 47.6-fold, corresponding to approximately 5.6 divisions. This predicts an average increase in cell number by more than 47-fold. Instead, the measured average increase was less than two-fold, suggesting a high level of cell death.
Figure 5. Summary of CFSE analysis of six samples. The CFSE signals of CD19+, CD10+ B cells (see Fig. 4A) at 1, 2, and 3 weeks of culture were normalized to those of CD19+, CD10– B cells. The latter signal is set at 100%. Numbers represent mean normalized signal, with standard deviation in parenthesis. Abbreviation: CFSE, carboxyfluorescein diacetate, succinimyl ester.
Growth Factors Increase Cell Division of Precursor B Cells
Previous studies demonstrated that growth factors can increase precursor B-cell expansion by up to 20-fold . In our cultures, the number of precursor B cells decreased 7-to 10-fold when growth factors were removed (n = 3). To determine how much of the decrease was the result of decreased cell division, CFSE-labeled CD19+ B cells were cultured with and without growth factors (Figs. 4A, 4B). Cultures were analyzed by flow cytometry at weeks 0, 1, 2, and 3. When cultured without growth factors, the CD19+, CD10+ precursor B cells showed continuous decrease in CFSE signal. However, the mean CFSE signal at 3 weeks was approximately two-fold more than that of B cells with growth factors, indicating that increased cell division contributed approximately only two-fold of the 7- to 10-fold increase in B cell number. The remainder of the increased number of B cells must be mediated by other processes. One possibility may involve apoptosis, because growth factors have been shown to promote survival in addition to cell division of precursor B cells .
Heterogeneous Rates of Cell Division among Cultured Leukemic Primary Precursor B Cells
Next, we asked if primary leukemic precursor B cells also had the same rate of cell division as the normal precursor B cells. We predicted that different leukemic cells would have different rates of cell division in culture, because previous studies on freshly isolated human precursor B cells indicated that leukemic B cells have variable rates of cell division . We studied precursor B leukemic cells from 10 patients (Table 1). No normal CD19+, CD10– B cells were present for normalization, preventing direct comparison between different samples. Instead, the signal was externally normalized to unlabeled Nalm-6, permitting comparison within a sample at different times. CD19+ B cells were isolated from excess bone marrow aspirates within 24 hours, labeled with CFSE, and cultured. At 0, 1, 2, and 3 weeks of culture, the cultures were trypsinized, stained with CD19 and CD10, and analyzed by three-color flow cytometry.
Table 1. Patient information and characteristic of the precursor B lymphoblastic leukemia
As predicted, the leukemic cells had variable rates of cell division (Fig. 6). Leukemic cells from patient A had a rapid decrease in CFSE signal of approximately 50-fold over 6 days (day 1 through day 7), corresponding to approximately 6.5 divisions per week. This division rate is more than three times that of normal precursor B cells. During the second and third weeks, the signal decreased by another 3.5- and 1.5-fold, respectively. However, these signals were near or at background and likely underestimate the number of cell divisions. Leukemic cells from patient B had a decreased CFSE signal of approximately 25-fold over 10 days (day 1 through day 11), corresponding to approximately 3.2 divisions per week. These leukemic cells also died rapidly in culture such that no viable cells were detected at 14 days of culture. Leukemic cells from patient C had a decreased CFSE signal of approximately 3.9-fold over 20 days (day 1 through day 21), corresponding to approximately 0.69 division per week, a rate approximately one third of that in normal precursor B cells. Leukemic cells from the other seven patients had cell division rates between the extreme rates of patient A and patient C.
Figure 6. CFSE analysis of three cases of precursor B lymphoblastic leukemia. Isolated CD19+ B cells from patients A, B, and C were labeled with CFSE, cultured, and analyzed as described in Figure 4. Abbreviations: CFSE, carboxyfluorescein diacetate, succinimyl ester; MFI, mean fluorescent intensity.
The slow rate of cell division of the leukemic cells from patient C was unexpected in a leukemia that is typically associated with an aggressive clinical behavior. Possible explanations for this unexpected finding included slowly dividing leukemic precursor B cells or culture conditions lacking an important proliferation signal present in vivo. To distinguish between these possibilities, we measured the proliferation index of the leukemic cells in the bone marrow biopsies from patients C and B. No biopsy was performed for patient A. The biopsies were placed in B5 fixative within 5 minutes of the procedure, ensuring rapid preservation of the cells near their original state. Examination of the H&E-stained sections of the biopsies revealed complete replacement of the marrow by leukemic cells (Figs. 7A, 7C). Proliferation cells were identified using paraffin immunohistochemical staining using antibodies against Ki67, a proliferation antigen with strong correlation to BrdU incorporation and mitotic rate . Biopsy of patient C showed that <5% of the leukemic cells were positive for Ki67 (Fig. 7B). In contrast, biopsy from patient B showed that >70% of the leukemic cells were positive for Ki67 (Fig. 7D). The biopsies from the other patients were analyzed similarly. The percentages of Ki67-positive cells were plotted against rate of CFSE-decreasing intensity for the nine leukemic samples with biopsies (Fig. 8). There was a general trend in which leukemic cells with high percentage of Ki67+ cells tended to have higher rates of cell division. These findings suggest that the cell division rates in culture for many primary leukemic cells correlate well with the in vivo rates.
Figure 7. Examination of bone marrow biopsy. (A, B): Patient C. H&E-stained section (A) and paraffin immunohistochemical-stained section with antibodies against Ki67 (B), magnification x400. Note the single strongly Ki67-positive cell in the lower left corner of (B). (C, D): Patient B. H&E-stained section (C) and paraffin immunohistochemical-stained section with antibodies against Ki67 (D), magnification x400. Abbreviation: H&E, hematoxylin and eosin.
Figure 8. Scatter plot of nine leukemic samples by percentage Ki67-positive versus in vitro cell division rate (decrease in CFSE signal per week). These represent patients 2 through 10 of Table 1. No biopsy was available for analysis for the first patient (patient A). Abbreviation: CFSE, carboxyfluorescein diacetate, succinimyl ester.
DISCUSSION
We thank Dr. Sindhu Cherian for her help in obtaining pathology information.
REFERENCES
Manabe A, Coustan-Smith E, Behm FG et al. Bone marrow-derived stromal cells prevent apoptotic cell death in B-lineage acute lymphoblastic leukemia. Blood 1992;79:2370–2377.
Pribyl JA, LeBien TW. Interleukin 7 independent development of human B cells. Proc Natl Acad Sci USA 1996;93: 10348–10353.
Moreau I, Duvert V, Banchereau J et al. Culture of human fetal B-cell precursors on bone marrow stroma maintains highly proliferative CD20dim cells. Blood 1993;81:1170–1178.
Barker J, Verfaillie CM. A novel in vitro model of early human adult B lymphopoiesis that allows proliferation of pro-B cells and differentiation to mature B lymphocytes. Leukemia 2000;14:1614–1620.
Fluckiger AC, Sanz E, Garcia-Lloret M et al. In vitro reconstitution of human B-cell ontogeny: from CD34+ multipotent progenitors to Ig-secreting cells. Blood 1998;92:4509–4520.
Nishigaki H, Ito C, Manabe A et al. Prevalence and growth characteristics of malignant stem cells in B-lineage acute lymphoblastic leukemia. Blood 1997;89:3735–3744.
Manabe A, Murti KG, Coustan-Smith E et al. Adhesion-dependent survival of normal and leukemic human B lymphoblasts on bone marrow stromal cells. Blood 1994;83: 758–766.
Mudry RE, Fortney JE, York T et al. Stromal cells regulate survival of B-lineage leukemic cells during chemotherapy. Blood 2000;96:1926–1932.
Ryan DH, Nuccie BL, Abboud CN. Inhibition of human bone marrow lymphoid progenitor colonies by antibodies to VLA integrins. J Immunol 1992;149:3759–3764.
Astier AL, Xu R, Svoboda M et al. Temporal gene expression profile of human precursor B leukemia cells induced by adhesion receptor: identification of pathways regulating B-cell survival. Blood 2003;101:1118–1127.
Kumagai M, Manabe A, Pui CH et al. Stroma-supported culture in childhood B-lineage acute lymphoblastic leukemia cells predicts treatment outcome. J Clin Invest 1996;97:755–760.
Kaaijk P, Kaspers GJ, Van Wering ER et al. Cell proliferation is related to in vitro drug resistance in childhood acute leukaemia. Br J Cancer 2003;88:775–781.
Nowicki M, Miskowiak B, Kaczmarek-Kanold M. Correlation between early treatment failure and Ki67 antigen expression in blast cells of children with acute lymphoblastic leukaemia before commencing treatment: a retrospective study. Oncology 2002;62:55–59.
Carter L, Fouser LA, Jussif J et al. PD-1:PD-L inhibitory pathway affects both CD4+ and CD8+ T cells and is overcome by IL-2. Eur J Immunol 2002;32:634–643.
Hasbold J, Lyons AB, Kehry MR et al. Cell division number regulates IgG1 and IgE switching of B cells following stimulation by CD40 ligand and IL-4. Eur J Immunol 1998; 28:1040–1051.
Agenes F. B lymphocyte life span, rate of division and differentiation are regulated by total cell number. Eur J Immunol 2003;33:1063–1069.
Lyons AB, Parish CR. Determination of lymphocyte division by flow cytometry. J Immunol Methods 1994;171: 131–137.
Wolf ML, Buckley JA, Goldfarb A et al. Development of a bone marrow culture for maintenance and growth of normal human B cell precursors. J Immunol 1991;147:3324–3330.
Lyons AB. Divided we stand: tracking cell proliferation with carboxyfluorescein diacetate succinimidyl ester. Immunol Cell Biol 1999;77:509–515.
Hirt A, Werren EM, Luethy AR et al. Cell cycle analysis in lymphoid neoplasia of childhood: differences among immunologic subtypes and similarities in the proliferation of normal and leukaemic precursor B cells. Br J Haematol 1992;80:189–193.
Ito M, Tsurusawa M, Zha Z et al. Cell proliferation in childhood acute leukemia: comparison of Ki-67 and proliferating cell nuclear antigen immunocytochemical and DNA flow cytometric analysis. Cancer 1992;69:2176–2182.
Birner P, Ritzi M, Musahl C et al. Immunohistochemical detection of cell growth fraction in formalin-fixed and paraffin-embedded murine tissue. Am J Pathol 2001;158: 1991–1996.
Young JL, Miller RW. Incidence of malignant tumors in U.S. children. J Pediatr 1975;86:254–258.
Markaki EA, Tsopanomichalou M, Dimitriou H et al. Mutations of retinoblastoma gene (Rb-1) as a prognostic factor in children with acute leukemia and neuroblastoma. Pediatr Hematol Oncol 2001;18:101–110.
Okuda T, Shurtleff SA, Valentine MB et al. Frequent deletion of p16INK4a/MTS1 and p15INK4b/MTS2 in pediatric acute lymphoblastic leukemia. Blood 1995;85:2321–2330.
Maloney KW, McGavran L, Odom LF et al. Different patterns of homozygous p16INK4A and p15INK4B deletions in childhood acute lymphoblastic leukemias containing distinct E2A translocations. Leukemia 1998;12:1417–1421.
Stock W, Tsai T, Golden C et al. Cell cycle regulatory gene abnormalities are important determinants of leukemogenesis and disease biology in adult acute lymphoblastic leukemia. Blood 2000;95:2364–2371.
Roman-Gomez J, Castillejo JA, Jimenez A et al. 5' CpG island hypermethylation is associated with transcriptional silencing of the p21(CIP1/WAF1/SDI1) gene and confers poor prognosis in acute lymphoblastic leukemia. Blood 2002;99:2291–2296.
Shen L, Kondo Y, Issa JP et al. Lack of p21(CIP1) DNA methylation in acute lymphocytic leukemia. Blood 2002; 100:3432–3433.
Mekki Y, Catallo R, Bertrand Y et al. Enhanced expression of p16ink4a is associated with a poor prognosis in childhood acute lymphoblastic leukemia. Leukemia 1999;13: 181–189.
Tsurusawa M, Ito M, Zha Z et al. Cell-cycle–associated expressions of proliferating cell nuclear antigen and Ki-67 reactive antigen of bone marrow blast cells in childhood acute leukemia. Leukemia 1992;6:669–674.
Ashley DM, Bol SJ, Kannourakis G. Measurement of the growth parameters of precursor B-acute lymphoblastic leukaemic cells in co-culture with bone marrow stromal cells: detection of two cd10 positive populations with different proliferative capacities and survival. Leuk Res 1994;18: 37–48.(Jonathan Coopermana, Robe)
b Pathology and Laboratory Medicine, University of Pennsylvania, Pennsylvania, USA;
c Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA
Key Words. Cell division ? Precursor B cell ? Lymphoblastic leukemia ? Stroma-based culture
Correspondence: John Kim Choi, M.D., Ph.D., Children’s Hospital of Philadelphia, 802F ARC, 3516 Civic Center Boulevard, Philadelphia, Pennsylvania 19104, USA. Telephone: 215-590-7194; Fax: 215-573-0342; e-mail: jkchoi@mail.med.upenn.edu
ABSTRACT
Primary normal and leukemic human precursor B cells can be cultured on bone marrow stromal cell layer for at least 3–4 weeks . Studies using this model have provided important insights into the cell pathways that promote the expansion of human precursor B cells. For example, expansion of normal or leukemic human B cells does not require interleukin-7 (IL-7) , a cytokine necessary for the expansion of murine precursor B cells. The rate of expansion is dictated, in part, by a balance of apoptosis and cell division. Apoptosis in both normal and leukemic precursor B cells is inhibited by direct contact with stromal cells and is mediated by vascular cellular adhesion molecule-1, 4?1 integrins, and multiple antiapoptotic proteins . Although the rate of apoptosis is similar among different samples of normal human precursor B cells, primary leukemic precursor B cells have variable rates of apoptosis; decreased apoptotic rate in culture is an independent predictor of poor clinical outcome .
Cell division is also an important contributor to precursor B cell expansion and may occur at a high rate even without expansion if there is offsetting apoptosis and differentiation. Studies of freshly isolated bone marrow cells demonstrate that the rates of cell division are similar among normal human precursor B cells but variable among leukemic precursor B cells. Furthermore, low rates of cell division in leukemic cells may correlate with poor response to chemotherapy and poor clinical outcome . As with apoptosis, stroma-based cultures provide a potentially powerful model for studying cell pathways that regulate cell division of primary normal and leukemic human precursor B cells. However, cell division has not been well characterized in the stroma-based culture model, and it is unknown how well precursor B cells divide in culture. In this study, we followed cell division using the fluorescent dye carboxyfluorescein diacetate, succinimyl ester (CFSE). This approach has not been previously used to study human lymphocytes but has been used to quantify cell division in murine lymphocytes . We report that different patient samples of normal precursor B cells have a relatively constant rate of cell division that is similar to reported in vivo rates. Cultured primary human leukemic cells have more heterogeneous rates of cell division that correlate with their in vivo proliferation index. These findings indicate that stroma-based cultures and CFSE can be used to study cell division of cultured primary human precursor B cells.
MATERIALS AND METHODS
Isolation and Characterization of Primary Normal Human CD19+ B Cells
Progenitor B cells were isolated by Ficoll gradient followed by using antibodies against CD19. Flow cytometry analysis of the purified cells showed >95% of the cells were positive for the B cell–specific antigen CD79a; of these, 70%–100% were positive for the precursor B cell antigen CD10 (Fig. 1). The CD19+, CD10+ cells were negative for the more mature B cell antigen, surface immunoglobulin heavy chain (IgM), and surface immunoglobulin light chains. The remaining 0%–30% of the CD19-selected cells was more mature CD79a+, CD10–, IgM+ B cells. More than 98% of the CD19+ cells were viable by 7-AAD exclusion (data not shown).
Figure 1. Characterization of isolated CD19+ B cells. Mononuclear bone marrow cells, before and after immunomagnetic selection with CD19 antibodies, were stained with phycoerythrin-conjugated anti-CD79a and allophycocyanin-conjugated anti-CD10 and analyzed by flow cytometry.
The CD19+ B cells were cocultured on stromal cells in X-vivo 10 media supplemented with 5% FBS and the growth factors IL-7, SCF, IGF-1, and FLT-3, as previously described . The AFT024 murine stromal cell line was replaced by primary pediatric bone marrow stromal cells, because in our hands the irradiated AFT024 cells continued to proliferate and overgrew the culture after 2 weeks (data not shown). At 0, 1, 2, and 3 weeks, the cells were trypsinized, counted, and analyzed by flow cytometry for CD19 and CD10 expression. Studies using a leukemic precursor B cell line, Nalm-6, indicated that trypsinization up to 15 minutes had no effect on the detection of CD19 and CD10 (data not shown). The expansions of CD19+, CD10+ precursor B cells (n = 6) were variable among different samples, with the cell numbers being zero to four times the initial number of plated cells at the end of 3 weeks. Most samples increased in number during the first 2 weeks and then decreased in number during the third week (Fig. 2). The variable expansions among samples and the decreased numbers after 2 weeks in culture are similar to the findings reported by others .
Figure 2. Expansion curves of CD19+, CD10+ precursor B cells from six patients. Isolated CD19+ B cells were cultured on bone marrow stroma. After 1, 2, and 3 weeks of culture, precursor B cells were enumerated by multiplying the total number of cells by the percentage of CD19+, CD10+ cells that was determined by flow cytometry analysis. The numbers were normalized to the numbers at initial plating.
Expansion of Cultured B Cells Is Determined by a Balance of Cell Division and Apoptosis
Two models can explain the variable, and in some cases lack of, expansion of precursor B cells among different samples with prolonged culture. In the first model, human B cells in some samples divide very slowly or not at all during prolonged culture. This predicts that cultures with no increase or net decrease in B cells have no cell division. In the other model, B cells are dividing but do not expand in number because of concomitant apoptosis. This predicts that all cultures contain dividing B cells. To determine which model is more accurate, we analyzed our cultures for active DNA synthesis by BrdU incorporation. Three-week-old cultures that had a net loss in B cells were incubated with BrdU, and the incorporated BrdU was detected using fluorescein-conjugated monoclonal antibodies to BrdU. UV microscopy showed many positive nuclei (Fig. 3), indicating that even in cultures without obvious B cell expansion, there was active cell division. The dividing cells were likely B cells, because the stromal cells were irradiated and thus unlikely to divide. In support, 3-week-old cultures of irradiated stromal cells without B cells demonstrated no BrdU incorporation (data not shown).
Figure 3. DNA synthesis and apoptosis in culture. Three-week-old cultures were incubated without (top row) and with (middle row) BrdU for 16 hours, stained with anti-BrdU antibodies, and examined under phase and UV microscopy (magnification x200). Three-week-old cultures were analyzed by terminal deoxynucleotidyl transferase–mediated rhodamine-dUTP nick-end labeling (TUNEL) and examined under phase and UV microscopy (magnification x 200).
Because there was a net decrease in B-cell number in some cultures despite DNA synthesis, apoptosis was likely present in our culture. To confirm this, 3-week-old cultures were examined for DNA fragmentation by TdT-mediated rhodamine-dUTP nick-end labeling (TUNEL). UV microscopy showed many positive nuclei (Fig. 3), confirming the presence of apoptosis. Three-week-old cultures of irradiated stromal cells without B cells showed virtually no TUNEL activity, suggesting that the dying cells were of B-cell origin (data not shown). These studies indicate that B-cell expansion in culture is dictated by a balance of cell division and apoptosis.
Rates of Cell Division in Different Samples of Cultured Normal Human B Cells Are Similar
Next, we asked if the variable expansions of precursor B cells could be explained by different rates of cell division among different samples. To quantify the rate of cell division, isolated human CD19+ B cells were labeled with CFSE, cultured, and followed for decreasing CFSE signal. In this approach, CFSE is covalently bound to intracellular macromolecules and eventually produces a stable fluorescent signal for each cell. With each division, the resulting daughter cells have half of the CFSE signal . At 0, 1, 2, and 3 weeks of culture, the cells were trypsinized, stained with antibodies against CD19 and CD10, and analyzed by three-color flow cytometry (Fig. 4A). Our CFSE analysis of human precursor B cells did not produce individual fluorescent peaks as seen with CFSE analysis of murine peripheral T lymphocytes . The lack of individual peaks most likely represented the heterogeneous sizes of the precursor B-cell population leading to overlapping peaks. Backgating on CD19+, CD10+ B cells confirmed a wide size distribution as determined by forward scatter; in contrast, the more mature CD19+, CD10– B cells showed more uniform size (data not shown). Nevertheless, this assay permitted a good estimation of the number of cell divisions.
Figure 4. CFSE analysis of normal human precursor B cells. (A): Isolated CD19+ B cells were labeled with CFSE and cultured on bone marrow stroma. After 4, 7, 14, and 21 days of culture, cells were stained with phycoerythrin-conjugated anti-CD19 and allophycocyanin-conjugated anti-CD10. Cells in the blast and lymphocyte region by forward and side scatter were subgated into CD19+, CD10+ and CD19+, CD10– B cells and then analyzed for CFSE expression. Figure is representative of six samples. (B): Same sample as in (A), except the growth factors were removed. Figure is representative of three samples. Abbreviations: CFSE, carboxyfluorescein diacetate, succinimyl ester; MFI, mean fluorescent intensity
Under our culture conditions, the more mature CD19+, CD10– B cells had a stable CFSE signal that decreased by only 25% over the 3 weeks of culture, suggesting little if any cell division within this population. The constant CFSE signal in this population confirmed the stable nature of the CFSE signal during prolonged culture. The constant CFSE signal also argued against significant differentiation of the dividing precursor cells, because this also results in CD19+, CD10– B cells with decreased CFSE signal. Hence, most of the CD19+, CD10– cells after 3 weeks in culture represented the initial plated cells.
Unlike the more mature CD10– B cells, cultured CD19+, CD10+ precursor B cells showed continuous decrease in CFSE signal throughout the 3 weeks of culture, indicating continuous cell division throughout the culture (Fig. 4). The rate of decreasing CFSE signal was examined for six different samples at weekly intervals (Fig. 5). The CFSE signals were normalized to the stable signal of the CD19+, CD10– B cells. This normalization compensated for the inherent variations in CFSE labeling between samples and flow cytometry settings between weeks. The samples had similar rates of cell division despite the variability in the expansion. During week 1 of culture, the average CFSE signal decreased by 3.14-fold, corresponding to approximately 1.65 divisions. During week 2 of culture, the CFSE signal decreased further by 3.88-fold, corresponding to approximately 1.95 divisions. During week 3 of culture, the CFSE signal decreased further by 3.90-fold, corresponding to approximately 1.89 divisions. Over the 3 weeks of culture, the CFSE signal decreased by 47.6-fold, corresponding to approximately 5.6 divisions. This predicts an average increase in cell number by more than 47-fold. Instead, the measured average increase was less than two-fold, suggesting a high level of cell death.
Figure 5. Summary of CFSE analysis of six samples. The CFSE signals of CD19+, CD10+ B cells (see Fig. 4A) at 1, 2, and 3 weeks of culture were normalized to those of CD19+, CD10– B cells. The latter signal is set at 100%. Numbers represent mean normalized signal, with standard deviation in parenthesis. Abbreviation: CFSE, carboxyfluorescein diacetate, succinimyl ester.
Growth Factors Increase Cell Division of Precursor B Cells
Previous studies demonstrated that growth factors can increase precursor B-cell expansion by up to 20-fold . In our cultures, the number of precursor B cells decreased 7-to 10-fold when growth factors were removed (n = 3). To determine how much of the decrease was the result of decreased cell division, CFSE-labeled CD19+ B cells were cultured with and without growth factors (Figs. 4A, 4B). Cultures were analyzed by flow cytometry at weeks 0, 1, 2, and 3. When cultured without growth factors, the CD19+, CD10+ precursor B cells showed continuous decrease in CFSE signal. However, the mean CFSE signal at 3 weeks was approximately two-fold more than that of B cells with growth factors, indicating that increased cell division contributed approximately only two-fold of the 7- to 10-fold increase in B cell number. The remainder of the increased number of B cells must be mediated by other processes. One possibility may involve apoptosis, because growth factors have been shown to promote survival in addition to cell division of precursor B cells .
Heterogeneous Rates of Cell Division among Cultured Leukemic Primary Precursor B Cells
Next, we asked if primary leukemic precursor B cells also had the same rate of cell division as the normal precursor B cells. We predicted that different leukemic cells would have different rates of cell division in culture, because previous studies on freshly isolated human precursor B cells indicated that leukemic B cells have variable rates of cell division . We studied precursor B leukemic cells from 10 patients (Table 1). No normal CD19+, CD10– B cells were present for normalization, preventing direct comparison between different samples. Instead, the signal was externally normalized to unlabeled Nalm-6, permitting comparison within a sample at different times. CD19+ B cells were isolated from excess bone marrow aspirates within 24 hours, labeled with CFSE, and cultured. At 0, 1, 2, and 3 weeks of culture, the cultures were trypsinized, stained with CD19 and CD10, and analyzed by three-color flow cytometry.
Table 1. Patient information and characteristic of the precursor B lymphoblastic leukemia
As predicted, the leukemic cells had variable rates of cell division (Fig. 6). Leukemic cells from patient A had a rapid decrease in CFSE signal of approximately 50-fold over 6 days (day 1 through day 7), corresponding to approximately 6.5 divisions per week. This division rate is more than three times that of normal precursor B cells. During the second and third weeks, the signal decreased by another 3.5- and 1.5-fold, respectively. However, these signals were near or at background and likely underestimate the number of cell divisions. Leukemic cells from patient B had a decreased CFSE signal of approximately 25-fold over 10 days (day 1 through day 11), corresponding to approximately 3.2 divisions per week. These leukemic cells also died rapidly in culture such that no viable cells were detected at 14 days of culture. Leukemic cells from patient C had a decreased CFSE signal of approximately 3.9-fold over 20 days (day 1 through day 21), corresponding to approximately 0.69 division per week, a rate approximately one third of that in normal precursor B cells. Leukemic cells from the other seven patients had cell division rates between the extreme rates of patient A and patient C.
Figure 6. CFSE analysis of three cases of precursor B lymphoblastic leukemia. Isolated CD19+ B cells from patients A, B, and C were labeled with CFSE, cultured, and analyzed as described in Figure 4. Abbreviations: CFSE, carboxyfluorescein diacetate, succinimyl ester; MFI, mean fluorescent intensity.
The slow rate of cell division of the leukemic cells from patient C was unexpected in a leukemia that is typically associated with an aggressive clinical behavior. Possible explanations for this unexpected finding included slowly dividing leukemic precursor B cells or culture conditions lacking an important proliferation signal present in vivo. To distinguish between these possibilities, we measured the proliferation index of the leukemic cells in the bone marrow biopsies from patients C and B. No biopsy was performed for patient A. The biopsies were placed in B5 fixative within 5 minutes of the procedure, ensuring rapid preservation of the cells near their original state. Examination of the H&E-stained sections of the biopsies revealed complete replacement of the marrow by leukemic cells (Figs. 7A, 7C). Proliferation cells were identified using paraffin immunohistochemical staining using antibodies against Ki67, a proliferation antigen with strong correlation to BrdU incorporation and mitotic rate . Biopsy of patient C showed that <5% of the leukemic cells were positive for Ki67 (Fig. 7B). In contrast, biopsy from patient B showed that >70% of the leukemic cells were positive for Ki67 (Fig. 7D). The biopsies from the other patients were analyzed similarly. The percentages of Ki67-positive cells were plotted against rate of CFSE-decreasing intensity for the nine leukemic samples with biopsies (Fig. 8). There was a general trend in which leukemic cells with high percentage of Ki67+ cells tended to have higher rates of cell division. These findings suggest that the cell division rates in culture for many primary leukemic cells correlate well with the in vivo rates.
Figure 7. Examination of bone marrow biopsy. (A, B): Patient C. H&E-stained section (A) and paraffin immunohistochemical-stained section with antibodies against Ki67 (B), magnification x400. Note the single strongly Ki67-positive cell in the lower left corner of (B). (C, D): Patient B. H&E-stained section (C) and paraffin immunohistochemical-stained section with antibodies against Ki67 (D), magnification x400. Abbreviation: H&E, hematoxylin and eosin.
Figure 8. Scatter plot of nine leukemic samples by percentage Ki67-positive versus in vitro cell division rate (decrease in CFSE signal per week). These represent patients 2 through 10 of Table 1. No biopsy was available for analysis for the first patient (patient A). Abbreviation: CFSE, carboxyfluorescein diacetate, succinimyl ester.
DISCUSSION
We thank Dr. Sindhu Cherian for her help in obtaining pathology information.
REFERENCES
Manabe A, Coustan-Smith E, Behm FG et al. Bone marrow-derived stromal cells prevent apoptotic cell death in B-lineage acute lymphoblastic leukemia. Blood 1992;79:2370–2377.
Pribyl JA, LeBien TW. Interleukin 7 independent development of human B cells. Proc Natl Acad Sci USA 1996;93: 10348–10353.
Moreau I, Duvert V, Banchereau J et al. Culture of human fetal B-cell precursors on bone marrow stroma maintains highly proliferative CD20dim cells. Blood 1993;81:1170–1178.
Barker J, Verfaillie CM. A novel in vitro model of early human adult B lymphopoiesis that allows proliferation of pro-B cells and differentiation to mature B lymphocytes. Leukemia 2000;14:1614–1620.
Fluckiger AC, Sanz E, Garcia-Lloret M et al. In vitro reconstitution of human B-cell ontogeny: from CD34+ multipotent progenitors to Ig-secreting cells. Blood 1998;92:4509–4520.
Nishigaki H, Ito C, Manabe A et al. Prevalence and growth characteristics of malignant stem cells in B-lineage acute lymphoblastic leukemia. Blood 1997;89:3735–3744.
Manabe A, Murti KG, Coustan-Smith E et al. Adhesion-dependent survival of normal and leukemic human B lymphoblasts on bone marrow stromal cells. Blood 1994;83: 758–766.
Mudry RE, Fortney JE, York T et al. Stromal cells regulate survival of B-lineage leukemic cells during chemotherapy. Blood 2000;96:1926–1932.
Ryan DH, Nuccie BL, Abboud CN. Inhibition of human bone marrow lymphoid progenitor colonies by antibodies to VLA integrins. J Immunol 1992;149:3759–3764.
Astier AL, Xu R, Svoboda M et al. Temporal gene expression profile of human precursor B leukemia cells induced by adhesion receptor: identification of pathways regulating B-cell survival. Blood 2003;101:1118–1127.
Kumagai M, Manabe A, Pui CH et al. Stroma-supported culture in childhood B-lineage acute lymphoblastic leukemia cells predicts treatment outcome. J Clin Invest 1996;97:755–760.
Kaaijk P, Kaspers GJ, Van Wering ER et al. Cell proliferation is related to in vitro drug resistance in childhood acute leukaemia. Br J Cancer 2003;88:775–781.
Nowicki M, Miskowiak B, Kaczmarek-Kanold M. Correlation between early treatment failure and Ki67 antigen expression in blast cells of children with acute lymphoblastic leukaemia before commencing treatment: a retrospective study. Oncology 2002;62:55–59.
Carter L, Fouser LA, Jussif J et al. PD-1:PD-L inhibitory pathway affects both CD4+ and CD8+ T cells and is overcome by IL-2. Eur J Immunol 2002;32:634–643.
Hasbold J, Lyons AB, Kehry MR et al. Cell division number regulates IgG1 and IgE switching of B cells following stimulation by CD40 ligand and IL-4. Eur J Immunol 1998; 28:1040–1051.
Agenes F. B lymphocyte life span, rate of division and differentiation are regulated by total cell number. Eur J Immunol 2003;33:1063–1069.
Lyons AB, Parish CR. Determination of lymphocyte division by flow cytometry. J Immunol Methods 1994;171: 131–137.
Wolf ML, Buckley JA, Goldfarb A et al. Development of a bone marrow culture for maintenance and growth of normal human B cell precursors. J Immunol 1991;147:3324–3330.
Lyons AB. Divided we stand: tracking cell proliferation with carboxyfluorescein diacetate succinimidyl ester. Immunol Cell Biol 1999;77:509–515.
Hirt A, Werren EM, Luethy AR et al. Cell cycle analysis in lymphoid neoplasia of childhood: differences among immunologic subtypes and similarities in the proliferation of normal and leukaemic precursor B cells. Br J Haematol 1992;80:189–193.
Ito M, Tsurusawa M, Zha Z et al. Cell proliferation in childhood acute leukemia: comparison of Ki-67 and proliferating cell nuclear antigen immunocytochemical and DNA flow cytometric analysis. Cancer 1992;69:2176–2182.
Birner P, Ritzi M, Musahl C et al. Immunohistochemical detection of cell growth fraction in formalin-fixed and paraffin-embedded murine tissue. Am J Pathol 2001;158: 1991–1996.
Young JL, Miller RW. Incidence of malignant tumors in U.S. children. J Pediatr 1975;86:254–258.
Markaki EA, Tsopanomichalou M, Dimitriou H et al. Mutations of retinoblastoma gene (Rb-1) as a prognostic factor in children with acute leukemia and neuroblastoma. Pediatr Hematol Oncol 2001;18:101–110.
Okuda T, Shurtleff SA, Valentine MB et al. Frequent deletion of p16INK4a/MTS1 and p15INK4b/MTS2 in pediatric acute lymphoblastic leukemia. Blood 1995;85:2321–2330.
Maloney KW, McGavran L, Odom LF et al. Different patterns of homozygous p16INK4A and p15INK4B deletions in childhood acute lymphoblastic leukemias containing distinct E2A translocations. Leukemia 1998;12:1417–1421.
Stock W, Tsai T, Golden C et al. Cell cycle regulatory gene abnormalities are important determinants of leukemogenesis and disease biology in adult acute lymphoblastic leukemia. Blood 2000;95:2364–2371.
Roman-Gomez J, Castillejo JA, Jimenez A et al. 5' CpG island hypermethylation is associated with transcriptional silencing of the p21(CIP1/WAF1/SDI1) gene and confers poor prognosis in acute lymphoblastic leukemia. Blood 2002;99:2291–2296.
Shen L, Kondo Y, Issa JP et al. Lack of p21(CIP1) DNA methylation in acute lymphocytic leukemia. Blood 2002; 100:3432–3433.
Mekki Y, Catallo R, Bertrand Y et al. Enhanced expression of p16ink4a is associated with a poor prognosis in childhood acute lymphoblastic leukemia. Leukemia 1999;13: 181–189.
Tsurusawa M, Ito M, Zha Z et al. Cell-cycle–associated expressions of proliferating cell nuclear antigen and Ki-67 reactive antigen of bone marrow blast cells in childhood acute leukemia. Leukemia 1992;6:669–674.
Ashley DM, Bol SJ, Kannourakis G. Measurement of the growth parameters of precursor B-acute lymphoblastic leukaemic cells in co-culture with bone marrow stromal cells: detection of two cd10 positive populations with different proliferative capacities and survival. Leuk Res 1994;18: 37–48.(Jonathan Coopermana, Robe)