a leukemic myeloid cell line with CD34+ progenitor and

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only.

1992 80: 1026-1032

OMA-AML-1: a leukemic myeloid cell line with CD34+ progenitor and CD15+ spontaneously differentiating cell compartments SJ Pirruccello, JD Jackson, MS Lang, J DeBoer, S Mann, D Crouse, WP Vaughan, KA Dicke and JG Sharp

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only.

OMA-AML-1: A Leukemic Myeloid Cell Line With CD34+ Progenitor and CD15+ Spontaneously Differentiating Cell Compartments By Samuel J. Pirruccello, John D. Jackson, Molly S. Lang, Joanne DeBoer, Sally Mann, David Crouse, William P. Vaughan, Karel A. Dicke, and J. Graham Sharp OMA-AML-1 was established from a patient with acute myelomonocytic (M4) leukemia at f i h relapse when blasts were greater than 85% CD34+, CD15-. Leukemic cells were established in suspension culture and independently grown as subcutaneous tumors in SCID mice. Cells growing in suspension culture underwent differentiation by phenotypic and morphologic criteria. In contrast, cells grown as subcutaneous solid tumors in SCID mice maintained progenitor cell characteristics with high-density CD34 expression and lack of morphologic differentiation.A tendency toward differentiation to CD15+, CD34- cells in vitro and self-renewal of CD34+, CD15- cells in vivo was consistently demonstrated regardless of whether cells were initially grown in vitro or in

vivo. The cell line maintains both a CD34+, CD15- progentitor cell pool and a non-overlapping, CD15+, CD34- differentiating cell compartment after more than 1 year in continuous culture. Cell cycle analysis and cloning experiments were consistent with terminal differentiation occurring in the CD15+,CD34- population. The cell line shows concentrationdependent proliferative responses to interleukin (IL)-3, granulocyte-macrophagecolony-stimulatingfactor (GM-CSF), and IL-6, but not to granulocyte CSF (G-CSF). OMA-AML-1 appears to mimic several features of normalmyeloid hematopoiesis and should prove useful for the study of normal and malignant myeloid differentiation. 0 1992 by The American Society of Hematology.

A

K562 can also be selectively induced to express cytoplasmic hemoglobin or to acquire megakaryocytic properties with appropriate These types of model systems have provided relatively homogenous cell populations for studying changes in gene expression and regulatory events that are otherwise difficult to study in the complex environment of long-term bone marrow cultures. We report here the initial characterization of a factor-independent, myeloid leukemic cell line that maintains both a CD34+, CD15progenitor cell pool and a CD15+, CD34- spontaneously differentiating cell compartment after more than 1 year in continuous culture.

CUTE MYELOGENOUS leukemia (AML) may simplistically be thought of as malignant transformation of a myeloid lineage cell frozen at an early stage of myeloid differentiation. Such cells generally show preferential selfrenewal at the expense of differentiation. Although within some French-American-British (FAB) subtypes of AML (eg, M2 to M4 and M5b) transformed cells may show limited differentiation in vivo, they more generally present as an expansion of early blast forms. In contrast, AML blasts placed in long-term culture under a variety of growth conditions frequently undergo terminal differentiation with subsequent loss of the clonogenic cell These studies have demonstrated that although leukemic myeloid blasts do not preferentially differentiate in vivo, they often retain the ability to differentiate in vitro. Very little is currently understood concerning the mechanisms underlying this behavior. In contrast to cytokine- or stromal-dependent leukemic blasts, which differentiate under the conditions of longterm culture, factor-independent, continuously cultured, myeloid leukemic cell lines generally maintain an undifferentiated morphology and phenotype. Many of these cell lines can be induced to differentiate under the influence of exogenous agents and have been used to model normal myeloid differentiation pathways. The myelomonocytic cell line HUO, for example, can be influenced to selectively undergo myeloid or monocyte lineage differentiation under the influence of dimethylsulfoxide (DMSO), vitamin D3, and several other agents!-’ The erythroleukemia cell line From the Departments of Pathology and Microbiology, Cell Biology and Anatomy, and Intemal Medicine, University of Nebraska Medical Center, Omaha, NE. Submitted August 19,1991; accepted April 30, 1992. Address reprint requests to Samuel J. Pirruccello, MD, Department of Pathology and Microbiology, University of Nebraska Medical Center, 600 S 42nd St, Omaha, NE 68198. The publication costs of this article were defrayed in pari by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 I992 by The American Society of Hematology. 0006-4971/92/8004-0016$3.00/0 1026

MATERIALS AND METHODS O M - A M L - I . The OMA-AML-1 cell line was established from leukemic blasts obtained from a patient with acute myelomonocytic leukemia at fifth relapse. At presentation the peripheral white blood cell count was approximately 25 x lo6 cells/pL with 90% blasts. The leukemic blasts were isolated by Ficollhypaque density gradient centrifugation from heparinized peripheral blood. The isolated blasts were cultured at 2 x lo6cells/mL in RPMI 1640 culture media supplemented with 10% fetal calf serum (GIBCO, Long Island, NY), 10% horse serum, mol/L 2-mercaptoethanol, and mol/L hydrocortisone sodium succinate (RFlOH10 media) at 37°C in a 5% COz in air atmosphere. Cultures were initially demidepopulated every 3 to 4 days and are now maintained by splitting every 5 days. All procedures were performed according to protocols approved by the University of Nebraska Institutional Review Board. Mice. C.B-17 SCID and nude mice used for these studies were bred and maintained in a pathogen-free environment within Class-I1 laminar air flow safety hoods (Baker, Stanford, ME) in microisolator cages in the Animal Resource Facility. SCID mice were screened for the “leaky” phenotype by assaying for serum immunoglobulin by radial immunodiffusion (Chemicon, El Segundo, CA) at 6 weeks of age. Mice without detectable serum immunoglobulin served as recipients. As prophylaxis against Pneumocystis carinii pneumonia, SCID mice were maintained on trimethoprim-sulfamethoxazole (0.32 g/mL trimethoprim and 1.6 glmL sulfamethoxazole) (Biocraft, Elmwood, NJ) in their drinking water for 3 of every 7 days. The remainder of the week the mice were provided with sterile, acidified water. Tumor cell xenografiing, animal monitoring, and tumor processing. All SCID mice were exposed to 3 Gy whole body irradiation 1 to 3 days before tumor cell inoculation. Approximately one half of the Blood, VOI 80,NO4 (August 15), 1992: pp 1026-1032

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only. OMA-AML-1

nude mice (designated as treated) were additionally immunosuppressed by injection with 0.05 mL anti-asialo GM1 (Wako, Dallas, TX) on day -3, followed by 4 Gy whole body irradiation on day -1 relative to transplantation. At the time cultures were initiated, 5 X lo7freshly isolated blasts were also inoculated into two SCID mice by intraperitoneal injection. In later experiments, mice were inoculated with 1 to 10 X lo7 SCID passaged or continuously cultured tumor cells by subcutaneous injection into the nape of the neck (SCID) or the flank (nude). The mice were monitored three times weekly for evidence of solid tumor outgrowth, and serial tumor size measurements were performed using calipers when tumors became evident. When tumors measured approximately 1.0 to 1.5 cm in diameter, mice were killed and the tumors were harvested. Tumors were bisected; one portion was placed into neutral buffered formalin for microscopic evaluation and the second portion was teased into RPMI 1640 to obtain a single-cell suspension. Single-cell suspensions were evaluated on WrightGiemsa-stained cytospin preparations and by flow cytometric phenotyping. At necropsy, the liver, spleen, heart, lungs, and kidneys were also removed, fixed in 10% buffered formalin, and processed for routine histological tissue sections and microscopic evaluation. The tumor and tissues were dehydrated and embedded in paraffin, cut at 5 pm, and stained with hematoxylin and eosin. Immunohistochemistry for Sudan black, myeloperoxidase, periodic acid-Schiff (PAS), and a-napthol acetate esterase (NSE) were performed on freshly isolated, cultured, and nude mouse xenotransplanted tumor cells by standard techniques. All animal protocols were performed in accordance with the guidelines of the University of Nebraska Medical Center Animal Review Committee. Cytogenetics. Cytogenetic analysis of freshly isolated and cultured leukemic blasts was performed using standard techniques.’* Flow cytomeby. Flow cytometric evaluation was performed on freshly isolated, cultured, and xenotransplanted tumor cell preparations according to previously described methods.13 Briefly, cells were aliquoted at 0.5 to 1 x lo6 cells/tube in 100 p L phosphatebuffered saline (PBS) containing 5% fetal bovine serum and 0.01% sodium azide. For single-color analysis, individual aliquots were incubated for 30 minutes at 4°C with saturating concentrations of primary antibody washed, and then incubated with goat anti-mouse fluorescein isothiocyanate (FITC). For two-parameter analysis, cells were incubated with biotin-conjugated anti-CD34 and FITCconjugated anti-CD1S antibodies, followed by incubation with streptavidin-phycoerythrin (PE). Background immunofluorescence was determined by staining an aliquot of cells with class- and conjugate-matched myeloma control antibodies, followed by appropriate secondary reagents. Cells were analyzed for percent positivity and mean channel fluorescence following background subtraction on a Coulter EPICS C or Profile I1 flow cytometer (Coulter, Hialeah, FL) using Coulter software. For two-parameter analysis of surface antigen expression and DNA cell cycle distribution, 5 x lo6 cells were first stained with anti-CD15 or anti-CD34, followed by goat anti-mouse FITC. The cells were gently fixed by incubating for 15 minutes in 0.5% paraformaldehyde in PBS and then permeabilized by a 5-minute incubation in 0.1% Triton X-100 in PBS. Following washing in PBS, lo6 cells were incubated for 1 hour in modified Vindelov’s solution containing 0.01% Nonidet P-4O.I4 The cells were then analyzed using the Coulter Profile I1 flow cytometer. Antibodies used for these studies included anti-HPCA-1 (CD34), anti-Leu-M1 (CD15), anti-Leu-MS (CDllc), anti-Leu-M1-FITC and streptavidin-PE (Becton Dickinson, Mountain View, CA); MY7 (CD13), MY4 (CD14), My9 (CD33), and myeloma control antibodies (Coulter); Ab-1-Biotin (CD34) (Oncogene Sciences, Manhasset, NY); and goat anti-mouse FITC (Tago, Burlingame, CA).

1027

Cell cloning. OMA-AML-1 cells were cloned in 100 p L of media at 1, 5, and 10 cells/well by limiting dilution into 96-well plates. Before cloning, cells were diluted with conditioned media from high-density cultures of OMA-AML-1 cells and two 96-well plates were plated at each cell density. Separately, OMA-AML-1 cells were stained with anti-CD34 or anti-CD1S antibodies and sorted on a Becton Dickinson FACStar Plus flow cytometer at 1,5, and 10 cells/well into two 96-well plates at each cell density. The plates were incubated for 2 weeks at 37°C in a humidified incubator containing 5% CO2 in air. The plates were fed with 50 p L of media at 1week postplating. At 2 weeks postplating, individual wells were read visually for cell proliferation under an inverted microscope. Wells were scored positive for cell proliferation as follows: 5 or more cells/well after plating at 1 cell/well, 20 or more cells/well after plating at 5 cells/well, and 40 or more cells/well after plating at 10 cells/well. Cytokine proliferation assay. Human recombinant (r)IL-3 (1 x lo8 U/mg protein), rGM-CSF (4 x lo7 Ulmg protein), and rIL-6 (3.0 ng/mL) were purchased from R&D Systems (Minneapolis, MN). OMA-AML-1 cells were plated in 96-well, flat-bottom microtiter plates at 5 x 103 cells/well. Each well contained a final volume of 0.1 mLmedium, which consisted of RPMI 1640,3% fetal calf serum (FCS), 100 U/mL penicillin, 100 pg/mL streptomycin, and doubling serial dilutions of the various cytokines. Background control wells contained media with no cytokines added. The initial concentrations of the cytokines added to the first well of the dilution titration were 2,000 U/mL IL-3, 2,000 U/mL GM-CSF, 2,000 U/mL G-CSF, and 500 ng/mL IL-6. The plates were incubated for 72 hours at 37°C in a humidified incubator containing 5% COz in air. At the end of the incubation period, a colorimetric microtiter tray (Ml’T) assay as described by M o s ~ m a n was ’ ~ used to measure the remaining cell viability. The mean of the O D value at 570 nm from the background control wells was subtracted from the O D values €or the cytokine-stimulated wells to obtain the plotted values. RESULTS

Establishment and characterization of the Om-AML-1 cell line. At the time of presentation, the leukemic blasts displayed morphologic and phenotypic characteristics consistent with an early myeloid cell. Morphologically, the blasts were relatively uniform in appearance with scanty cytoplasm and a high nuclear to cytoplasmic ratio (not shown). Some blasts displayed horseshoe-shaped nuclei. Primary granules were also evident in some cells. Immunohistochemistry demonstrated weak positivity for Sudan black and NSE, indicating monocytic differentiation. Cytogenetic studies showed a diploid tumor with a complex translocationinvolving chromosomes 2,7, and 11 [46,XY,t(2; 7;11)(p21;p12;p15)]. The majority of blasts were positive for HLA-DR, the pan-myeloid antigens CD13 and CD33, and the stem cell antigen CD34 (Table 1, preculture). A smaller percentage of blasts were also expressing the more mature myeloid antigens C D l l c and CD15. The blasts were noted to be expressing relatively high CD34 antigen density based on mean channel fluorescence (Fig 1A). The freshly isolated leukemic blasts were established in Dexter-type culture media (RFlOH10) in the absence of a preestablished stromal cell layer. Morphologic assessment, phenotyping, immunohistochemistry, and cytogenetic studies were repeated at 1- to 3-month intervals during the early postculture period and at longer intervals after the first 6

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only. PIRRUCCELLO ET AL

1028 Table 1. Phenotypic Profile of OMA-AML-1 In Vivo and In Vitro Postculture'

SCID

HLA-DR CD1l c CD13 CD14 CD15 CD33 CD34

Preculture

Mouse Tumor

3 mo

1 yr

57.lt 45.7 88.4 2.8 22.0 87.6 07.4

22.6 ND 24.1 4.6 3.0 96.1 92.9

4.0 34.1 52.0 27.0 44.8 71.4 17.0

0.2 51.0 56.8 2.2 76.5 95.2 20.5

Abbreviation: ND, not determined. 'Cells continuously cultured in RF10H10 media. tPercentage of cells positive for antigen.

months in culture. In contrast to the preculture studies, less than 40% of the cultured cells had characteristic blast morphology. Greater than 50% of the cultured cells displayed monocyte morphology or multilobulated nuclei with abundant cytoplasm and light granulation (myeloid differentiation) (Fig 2A). Differential count of cells continuously cultured for more than 5 months yielded 40% monocytes, 10% maturing myeloid cells, 10% neutrophils, 5% granulated blasts, and 35% myeloblasts. Immunohistochemistry demonstra&ed20% of the cultured cells to be positive for NSE, which was still weakly positive after fluoride inhibition. We could not demonstrate reduction of nitroblue tetrazolium dye by the cultured cells; however, the monocytes that appeared to be the most differentiated were phagocytic for Staphlycoccus aureus. Cytogenetic analysis demonstrated the complex translocation t(2;7;1l)(p21;p12; p15) and the acquisition of an unidentified marker chromosome. In comparison to the preculture phenotype, the cultured cells showed a much smaller percentage of CD34+ cells and a considerable increase in CD15+ cells, consistent with the morphologic differentiation observed (Table 1, postculture;

Fig 1B). In the early postculture studies, an increase in CD14+ cells was also noted; however, after 1 year in continuous culture, only a small percentage of the cultured cells continued to express this monocyte lineage antigen (Table 1, postculture, 3 months and 1year). Differentiation in vitro versusprogenitor self-renewalin vivo. At the time cultures were initiated, 5 X lo7 freshly isolated blasts were transferred by intraperitoneal injection into two irradiated SCID mice. The tumor cells did not grow in the peritoneal cavity. However, approximately 6 weeks after cell transfer, a subcutaneous tumor was noted near the site of injection in one mouse of the pair. Tumor cells were harvested for study 8 weeks posttransfer, when the tumor was approximately 1.0 cm in diameter. The cells comprising the tumor primarily displayed an undifferentiated blast-like morphology with none of the differentiation characteristics of the cultured cells (Fig 2B and C). Phenotypically, blasts were greater than 90%, and had high-density CD34 antigen expression with less than 5% of the cells expressing the later myeloid lineage antigen CD15 (Table 1, SCID mouse; Fig IC). Approximately lo7 tumor cells isolated from the SCID mouse were reinoculated subcutaneously into two more SCID mice. A second aliquot of cells was placed in suspension culture in RFlOHlO media. After repassage in the SCID mouse, the tumor cells continued to demonstrate an undifferentiated morphology and were greater than 90% CD34+ and less than 5% CD15+. In contrast, the cells placed in culture demonstrated both morphologic and phenotypic evidence of differentiation with less than 20% CD34+ cells and greater than 60% CD15+ cells. When the SCID to culture adapted cells were repassaged subcutaneously in a second SCID mouse, 95% of the harvested cells were morphologically blast cells and all of these cells expressed high-density CD34. In another series of experiments, the original cultured

B

A Control

COnUol

CD'5

CDH

cm

Log Green Fluorescence

I I Fig 1. Log scale fluorescence histogram profiles of CD34+ and CD15+ cell distributionsdemonstrating differentiation in vitro and self-renewal in vivo. (A) Freshly isolated leukemic blasts; (E) Cells derived from long-term continuous culture; (C) xenotransplantedtumor cells from nude mouse host.

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only. OMA-AML-1

1029

Fig 2. Photomicrographs of OMA-AML-1 cells demonstrating differentiation in vitro and self-renewal in vivo. (A) Wright-Giemsa-rtained cytospin prepation of cells from long-term continuous culture. Differentiated cells with multilobulated nuclei are abundant. (Original magnification x 1.000.) (6)Wright-Geimsa-stained cytospin preparation of leukemic cells isolated from SCID mouse tumor. The majority of cells show a blast-like morphology. (Original magnification x 1.000.) (C) Hematoxylin and eosin-stained section of SCID mouse tumor. The blasts are monotonous in appearance and numerous mitotic figures are evident. (Original magnification x200.)

cells were subcutaneously passaged in SCID and nude mice. The cultured cells demonstrated growth characteristics similar to the original SCID tumor in both untreated and treated (ie, irradiation plus anti-asialo GMl) nude and SCID hosts (Fig 3). The suspension cultured cells recov-

Fig 3. Composite growth curves of OMAAML-1 cells xenotransplanted t o (W) treated (irradiation and anti-asialo GM1) and ( 0 ) untreated nude mice. Each group represents the average of five mice. Tumor engraftment occurred in three of five treated mice and four of five untreated mice. There were no significant differences in tumor growth patterns in the untreated and additionally immunosuppressed mice.

cred from cithcr SCID or nude hosts showed a dramatic increase in the percentage of CD34+ cells ( > 60%) and a concomitant decrease in CD15+ cells. When tumor cells harvested from a nude host were reintroduced to culture, the cells rapidly reexpressed the differentiated morphology and phenotype. One week after reintroduction to culture, greater than 75% of the cells expressed CD15. Cultured cells maintain distinct CD34+,CD15-, progentitor cell and CD15+,CD34-, terminallydifferentiating cellpopirla(ions. After more than 1 year in culture, the OMA-AML-I cell line appeared stable with respect to the rclativc distributions of CD34+ and CD15+ cells. Generally, the percentage of cells positive for these two markers, together, accountcd for 100% of the cells analyzed. We had prcviously noted that the cells with the highest forward and 90" light scatter properties also showed fewer CD34+ cells and a higher percentage of cells expressing the differentiation antigens CD15, CDllc, and CD14 (not shown). To dctermine the exact relationship of the stem cell antigen (CD34)positive cells to the differentiation antigen (CD15)-positivc cells, two-parameter analysis was performed. This analysis demonstrated that the OMA-AML-1 cell line comprised two distinct populations of cells: a CD34+, CD15- population, and a CD15+,CD34- population (Fig 4). We reasoned that the CD34+, CD15- cells might represent a progenitor cell pool and should therefore exhibit cell cycle distribution characteristics different from the more differentiated CD15+, CD34- cells. Using surface antigen gating in combination with DNA cell cycle analysis, we


1030

Table 2. Cell Cycle Analysis of CD15+ Versus CD34+ OMAAML-1 Cells Cells Cultured

Cells Cultured

at 1-2 x 10@Cells/mL

GoGl S G2M

at

< 1 x 106 Cells/mL

CD15+

CD34+

Total

CD15+

CD34+

Total

85.3* 9.5 5.2

61.8 24.2 14.0

83.0 8.7 8.3

93.6 4.7 1.7

88.2 9.1 2.7

94.6 3.9 1.5

*Percentage of cells in cell cycle compartment.

IgG-FITC Fig 4. Two-parameter fluorescence histograms demonstrating that OMA-AML-1 comprises t w o distinct, non-overlapping CD34+ and CD15+ populations. Cells in continuous culture were stained with CD34-biotin and CD15-FITC, followed by streptavidin-PE, and then analyzed on a Profile II flow cytometer. The top histogram represents the background control.

W

4

iu3

CD15-FlTC

found that the CD34+ population had at least twice as many cells in S+G,M as the CD15+ population (Fig 5, Table 2). These differences in proliferative activity could be consistently demonstrated regardless of the initial cell culture densities and overall cell cycle distribution profile. We then performed a cloning experiment to determine whether the CD15+, CD34- cells represented a terminally differentiating cell compartment. Suspension-cultured

OMA-AML-1 cells were subcloned by limiting dilution into 96-well plates at 1,5,and 10 cells/well. Separately, CD15+, CD34- and CD34+, CD15- cells were sorted at 1,5, and 10 cells/well, and wells were monitored for a quadrupling of cell numbers at 2 weeks postplating. We found that the CD15+, CD34- cells showed no proliferation at 1, 5, or 10 cells/well by this criterion (Fig 6). In contrast, the CD34+, CD15- cells demonstrated greater than 30% of wells with cell quadrupling when plated at 10 cells/well. However, the best cell growth was demonstrated for the cells plated by limiting dilution, with greater than 60% of the wells plated at 10 cells/well showing growth. Om-AML-1cells respond to hematopoietic growthfactors. Although the OMA-AML-1 cells are growth factorindependent, the addition of certain cytokines enhanced their proliferative activity (Fig 7). IL-3 and GM-CSF increased the proliferation of the cell line in a dosedependent manner. At higher concentrations, IL-6 had no 70 1

I"

I

0,

C ._

3

0 L

v, Red Fluorescence (PI)

n Q, Y

-O

4

a

\c

0

w 1

5

10

Red Fluorescence (PI)

Number of Cells/Well Plated Fig 5. Fluorescence histograms of OMAAML-1 demonstrating increased proliferative (S G,M)-phase cells in the CD34+ population in comparison t o the CD15+ cells. The two-parameter histograms on the left show the ungated profiles for surface antigen expression (y-axis) and DNA content (x-axis). The CD34+ cells (top left) show significantly more cells in S G2Mthan the CD15+ cells (bottom left). The single-parameter histograms on the right represent the DNA cell cycle profiles following surface antigen gating for CD34+ cells (top right) and CD15+ cells (bottom right). Analysis windows used for calculating the cell cycle distribution fractions reported in Table 2 are illustrated.

+

+

Fig 6. OMA-AML-1 cell cloning assay demonstrating that terminal differentiation is occurring in the CD15+, CD34- cell population. Cells were plated at 1,5, or 10 cells/well into 96-well plates either by (V) limiting dilution or by sorting for ( 0 )CD34+ or (m) CD15+ cells. Individual wells were then evaluated for cell growth at 2 weeks postplating on an inverted microscope. Wells were scored as positive for cell growth if the plated cells had quadrupled in cell number. Each point represents the mean percentage of positive wells averaged for two 96-well plates. Note that cell proliferation was detected only in the wells containing CD34+cells.

From bloodjournal.hematologylibrary.org by guest on July 11, 2011. For personal use only. 1031

OMA-AML. 0.04

9

\

0.03

A

E

E 0

0.02

v

0

0 0.01

0.oc

1

2

3

i

5

6

7

8

9 1 0 1 1 1 2

1/Log2 Dilution Fig 7. MTT assay demonstrating concentration-dependentproliferative responses of OMA-AML-1 cells to specific cytokines. Cells were plated at 5 x l o 3 cells/well into 96-well plates in RPMl 1640 with 3% G-CSF were FCS. The cytokines ( 0 )IL-3, ( 0 )GM-CSF, (VIIL-6, or (0) then added as serial doubling dilutions. The initial concentrations of the cytokines added to the first well of the dilution titration were 2,000 U/mL IL-3.2.000 U/mL GM-CSF, 2,000 U/mL G-CSF, and 500 ng/mL IL-6. The plates were incubated for 72 hours at 37°C in a humidified incubator containing 5% CO, in air. At the end of the incubation period, the mean value of the OD at 570 nm from the background control wells (ie, no cytokines) was subtracted from the mean ODs for the cytokine-stimulatedwells to obtain the plotted OD values.

additional growth promoting effects; however, at lower concentrations, IL-6 increased the proliferation of OMAAML-1 cells. G-CSF had no effect on cell proliferation except at the highest concentration. DISCUSSION

The OMA-AML-1 cell line, established from a patient at fifth relapse with M4 AML, appears to be unique in comparison to other myeloid leukemic cell lines. The presentation and dynamics of the M4 FAB subtype of AML suggests that the clonogenic cell represents a common precursor for both the monocytic and myeloid lineages. In the patient from whom OMA-AML-1 was established, disease progression (each subsequent relapse) was associated with a loss of differentiation and expansion of a more undifferentiated, CD34+ myeloid blast cell population. The adaption of these blasts to suspension culture was associated with morphologic differentiation to intermediate- and end-stage monocytes and myeloid cells. The percentage of morphologically differentiated cells has remained relatively constant after more than 1 year in continuous culture and correlates well with the relative number of CD15+, CD34cells. The ability to grow cells obtained from the peripheral blood of this patient in suspension culture, as well as in SCID mice, indicates that the blasts were factor- and stromal cell-independent at the time of isolation. We have

attempted long-term culture and SCID mouse xenografting with several other AMLs with similar phenotypes with only limited success (unpublished observations, Pirruccello and Sharp). Many of these patient derived AMLs exhibit only differentiation in long-term bone marrow culture.1-3 In contrast to acute lymphoblastic leukemia (ALL) cell lines, which can be readily engrafted in SCIDs by intraperitoneal, intravenous, or subcutaneous routes,16-18the OMA-AML-1 cells only showed growth following subcutaneous inoculation. We were unable to engraft these cells by either intravenous or intraperitoneal inoculation. Further, the solid tumors formed by OMA-AML-1were poorly vascularized, often showed central areas of necrosis, and never metastasized to other organs. However, we did not attempt kidney capsule implantation with OMA-AML-1, which has been reported to be a successful technique for engrafting fresh patient AMLs in SCID mice.19 The ability to passage OMA-AML-1 subcutaneously in nude mice was unexpected, since nudes have generally been considered less-conducive hosts for growth of human tumors?O Of note, OMA-AML-1 cells showed similar growth curves in both untreated nude mice and nude mice pretreated with irradiation and anti-asialo GM1. Our own experience with xenografting ALL cell lines has also demonstrated nude mice to be suitable hosts with somewhat longer survival times than SCIDs inoculated with the same ALL cell dose.21 The fact that OMA-AML-1 preferentially differentiates in vitro while favoring progenitor cell expansion in vivo is of some interest. This was consistently demonstrated regardless of whether the cells were initially grown in vitro or in vivo. We noted that following xenografting of OMAAML-1 in the nude mice, tumor regression occurred over the first 3 to 5 days after cell transfer, followed by later tumor expansion. This early tumor regression may have represented loss by terminal differentiation of the CD15+, CD34- cell compartment. This would have then been accompanied by a loss of further differentiation of the CD34+, CD15- progentitor cell pool. Alternatively, it is possible that the differentiated cells were more susceptible than the progentitor cells to residual mouse natural killer cell or other cytotoxic effector cell function. However, the observed behavior is consistent with previous observations of acute myeloid leukemias in both humans and the BNML rat myeloid leukemia model where little differentiation is seen in vivo, yet terminal differentiation is favored in vitro.22 Most factor-independent, continuously cultured myeloid leukemia cell lines represent early myeloid precursors and are relatively homogenous phenotypically and morphologically. Although differentiation can be induced, these continuously cultured cell lines generally show little spontaneous differentiation and may show reversion to a more immature phenotype.23 The OMA-AML-1 cell line maintains both a CD34+, CD15- progenitor cell population and a nonoverlapping, CD15+, CD34- spontaneously differentiating population after more than 1 year in continuous culture. The CD34+ population generally comprises approximately 30% of the cultured cells; however, we have noted fluctuations in the relative distribution of the progenitor cell pool


1032

(ie, from 10% to 50% CD34+ cells) that appears to be unrelated to cell density. OMA-AML-1 demonstrates concentration-dependent proliferative responses to the early-acting cytokines IL-3, GM-CSF, and IL-6, but not to the late-acting factor G-CSF. This finding suggests that the proliferative response was occurring in the CD34+,CD15- progenitor cell population. This was further supported by the DNA cell cycle and cloning experiments, which indicate that terminal differentiation is occurring in the CD15+,CD34- cell compartment. The relative distributions of the progenitor and differentiating cell compartments can also be modulated by a number of cytokines such as IL-3 and leukemia inhibitory factor (LIF).24The latter cytokine induces a dramatic reduction in

the CD34+, CD15- population and a concomitant increase in the CD15+, CD34- population. In addition to the proliferative and phenotypic differentiative responses, we have also noted induction of eosinophil production in response to IL-3 and GM-CSF.” These findings indicate that OMA-AML-1 is multipotential and capable of relatively complex responses to various hematopoietic growth factors. The OMA-AML-1 cell line therefore appears to mimic some features of normal myelopoiesis by maintaining discreet progentitor cell and spontaneously differentiating cell compartments. This cell line should prove to be a valuable tool for investigating the dynamics of normal and malignant hematopoiesis.

REFERENCES

1. Schiro R, Coutinho LH, Will A, Chang J, Testa NG, Dexter TM: Growth of normal versus leukemic bone marrow cells in long term culture from acute lymphoblastic and myeloblastic leukemias. Blut 61:267,1990 2. Baer MR, Gao XZ, Preisler HD: Sustained viability of acute myelogenous leukemia cells in liquid suspension culture correlates with occurrence of differentiation. Leuk Res 12:647,1988 3. DeBoer J, Welniak L, Pirruccello S, Mann S, Vaughan W, Sanger W, Sharp JG: Cytogenetic evidence suggesting that both normal and leukemic stem cells differentiate in long term culture. Applied Cytogenet 17:62, 1991 (abstr) 4. Kharbanda SM, Sherman ML, Kufe DW: Effects of tiazofurin on guanine nucleotide binding regulatory proteins in HL-60 cells. Blood 75583,1990 5. Pedrazzoli P, Bains MA, Watson R, Fisher J, Hoy TG, Jacobs A. C-myc and c-myb oncoproteins during induced maturation of myeloid and erythroid human leukemic cell lines. Cancer Res 24:1, 1989 6. Ladoux A, Krawice I, Damais C, Frelin C: Phorbol esters and chemotactic factor induce distinct changes in cytoplasmic CaZ+and pH in granulocytic like HL60 cells. Biochim Biophys Acta 1013:55, 1989 7. Wang SY, Liu ST, Wang SJ, Ho C K Induction of differentiation in HL-60 cells by retinoic acid and lymphocyte-derived differentiation-inducing factor but not by recombinant G-CSF and GM-CSF. Leuk Res 13:1091,1989 8. Honma Y, Okabe KJ, Hozumi M, Uehara Y, Mizuno S: Induction of erythroid differentiation of K562 human leukemic cells by herbimycin A, an inhibitor of tyrosine kinase activity. Cancer Res 49:331,1989 9. Alitalo R, Partanen J, Pertovaara L, Holtta E, Sistonen L, Anderson L, Alitalo K Increased erythroid potentiating activity/ tissue inhibitor of metalloproteinases and jun/fos transcription factor complex characterize tumor promoter-induced megakaryoblastic differentiation of K562 leukemia cells. Blood 75:1974, 1990 10. Alitalo R: Induced differentiation of K562 leukemia cells: A model for studies of gene expression in early megakaryoblasts. Leuk Res 14501,1990 11. Yu J, Lemas V, Page T, Connor JD, Yu AL: Induction of erythroid differentiation in K562 cells by inhibitors of inosine monophosphate dehydrogenase. Cancer Res 49555,1989

12. Yunis JJ: High resolution of the human chromosomes. Science 191:1268,1976 13. Pirruccello SJ, Bicak MS, Gordon BG, Gajl PK, Gnarra DJ, Coccia PF: Acute lymphoblastic leukemia of NK-cell lineage: Responses to IL-2. Leuk Res 13:735,1989 14. Vindelov L L Flow microfluorometric analysis of nuclear DNA in cells from solid tumors and cell suspension. Virchows Arch [B] 24227,1977 15. Mossman T: Rapid colorimetric assay for cellular growth and survival: Application to proliferation and cytotoxicity assays. J Immunol Methods 65:55,1983 16. Kamel-Reid S, Letarte M, Sirard C, Doedens D, Grunberger T, Fulop MH, Freedman RA, Phillips RA, Dick JE: A model of human acute lymphoblastic leukemia in immunodeficient SCID mice. Science 246:1597,1989 17. Cesano A, O’Conner R, Lange B, Finan J, Rovera G, Santoli D: Homing and progression patterns of childhood acute lymphoblastic leukemias in severe combined immunodeficient mice. Blood 17:2463,1991 18. Pirruccello SJ, Lang MS, Kleveland KK, Sharp JG, Beisel KW, Dicke KA, Purtilo DT: Evaluation of the SCID mouse as host for growth of human hematopoietic tumors. Blood 76:309a, 1990 (abstr, suppl 1) 19. Sawyers CL, Gishizky ML, Quan S, Golde DW, Witte ON: Propagation of human blastic myeloid leukemias in the SCID mouse. Blood 79:2089,1992 20. Phillips RA, Jewett MAS, Gallie BL: Growth of human tumors in immune-deficient scid mice and nude mice. Curr Top Microbiol Immunol152259,1989 21. Lang M, Maloch M, Purtilo DT, Pirruccello SJ: Engraftment parameters of human pre-B ALL cell lines in SCID and nude mice. (in preparation) 22. Colly LP, van Bekkum DW, Hagenbeek A: Cell kinetic studies after high dose Ara-C and Adriamycin treatment in a slowly growing rat leukemia model (BNML) for human acute myelocytic leukemia. Leuk Res 8:945,1984 23. Leglise MC, Dent GA, Ayscue LH, Ross DW: Leukemic cell maturation: Phenotypic variability and oncogene expression in HL60 cells: A review. Blood Cells 13:319,1988 24. Jackson JD, Korth D, Lang MS, Pirruccello SJ: Cytokine induced differentiation of the myeloid leukemic cell line OMAAML-1. (in preparation)