From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
Characterization of Adhesion Molecules on Human Myeloma Cell Lines
By Hiroshi Uchiyama, Bruce A. Barut, Dharminder Chauhan, Stephen A. Cannistra, and Kenneth C. Anderson
In multiple myeloma, malignant plasma cells are localized in
marrow and rarely circulate in peripheral blood. To investigate the role of adhesion proteins in this process, we
determined the expression and function of adhesion molecules on cell lines derived from patients with myeloma. The
U266, ARH-77, IM-9, and HS-Sultan cell lines strongly expressed p1 and a4 integrins (89% t o 98% positive), confirming that VLA-4 is the principal integrin on these cell lines. The
U266 and IM-9 cell lines also expressed cs, integrin on 15% t o
A,and
20% cells. In contrast, all lines lacked cell surface d,
a6 integrin expression (<5% positive). These cell lines adhered t o fibronectin (20% t o 40% specific binding), without
significant binding t o either collagen or laminin. Adhesion of
these cell lines t o fibronectin was partially blocked with
either anti-pl integrin monoclonal antibody (MoAb) (75%
inhibition), a n t i - d integrin MoAb (75% inhibition), or RGD
peptide (50% inhibition), but was unaffected by anti-wp3 or
anti-allbp3 MoAbs. Moreover, the combination of anti-61
plus RGD peptide or a n t i - d plus RGD peptide inhibited
binding t o fibronectin by 80% and 95%, respectively. Finally,
pretreatment and coculture of the IM-9 cell line with interleukin-6 (IL-6) resulted in a 52% decrease in specific binding t o
fibronectin (30% k 6% t o 15% f 6%; P = .001), associated
with a decrease in the number of cells expressing VLA-4 and
a decrease in intensity of VLA-4 expression. These data
suggest that myeloma cells adhere t o fibronectin through
VLA-4 as well as through RGD-dependent mechanisms, and
that this binding can be downregulated by IL-6. Future
studies of binding of both myeloma cell lines and freshly
isolated tumor cells t o extracellular matrix proteins and t o
marrow stroma may enhance our understanding of localization and trafficking of cells within the bone marrow microenvironment.
o 1992b y The American Society of Hematology.
I
pl(CD29), P2(CD18) and P3(CD61) integrins? Members
of the p l integrin subfamily (also called VLA proteins)
contain the p l subunit in association with one of at least six
different a subunits and include receptors that bind to the
ECMs fibronectin, laminin, and collagen. VLAl(alp1) and
VLA2(a2pl) are the receptors to collagen and laminin;
VLA3(a3pl) mediates binding to fibronectin and laminin,
VLA4(a4pl) and VLAS(aSp1) bind to fibronectin; and
VLA6(a6pl) adheres to laminin6 (Table 1).
Preliminary studies of marrow myeloma cells suggest that
they lack a2p1, a3p1, or a 6 p l and have only variable
expression of a4pl integrin~.~
The lack of a 2 p l and a 6 p l
expression is consistent with their inability to invade endothelial basement membranes, which are composed of collagen type IV and laminin. LFA-1 is a p2 integrin that binds
to its receptors, ie, intercellular adhesion molecule
(1CAM)-1 and ICAM-2.R39LFA-1/ICAM interactions are
required for the induction of a broad spectrum of leukocyte
functions such as cytotoxicity,1° proliferation,ll and differentiation,12as well as lymphocyte binding to activated endothelial cells in migration and h ~ m i n g . ’ ~Although
.’~
normal
plasma cells lack LFA-1, expression of LFA-1 on malignant
plasma cells is correlated with proliferation, suggesting that
cellular contact is critical for proliferation of malignant
plasma cells in myeloma.l5Finally, the strong expression of
the lymphocyte homing receptor CD44 on the majority of
myelomas, plasma cell leukemias, and myeloma cell lines is
of particular interest because CD44 has been linked to
site-specific extravasation of lymphocytes into tissues, sitedirected homing capabilities in B and T lymphocytes, and
binding to ECMs.16 Studies of these and other adhesion
molecules on malignant plasma cells may help to gain
insight into processes of growth, spread, and clinical behavior of multiple myeloma.
In the present study, we attempted to identify the
adhesion molecules on the surface of five myeloma-derived
cell lines and to examine their role in adhesion to ECMs.
On the U266, ARH-77, IM-9, and HS-Sultan human
myeloma-derived cell lines, VLA-4 is the principal integrin
expressed; these cell lines adhered to fibronectin, with little
or no binding to either collagen or laminin, consistent with
N MOST CASES of multiple myeloma, the malignant
plasma cells or plasmablasts are localized in the bone
marrow (BM) and rarely circulate in the peripheral blood.
Antiidiotypic antibody, as well as phenotypic studies, confirm that malignant pre-B and B cells can be found in
peripheral blood in the setting of myel~mal-~;
however,
circulating plasma cells occur only rarely, with the late
development of plasma cell leukemia. Preliminary studies
suggest that myeloma cells cannot bind to rat highendothelial venules (HEV), consistent with their lack of
migration into peripheral blood.4 In contrast, plasma cell
leukemia cells do adhere to rat HEV, facilitating their exit
into the circulation. Nonetheless, the mechanisms by which
myeloma cells localize to the BM, via binding either to
extracellular matrix proteins (ECMs) or to myeloma marrow stroma, remain undefined.
Several adhesion molecules have been defined that
mediate cellular interactions and cellular migration. The
known cellular adhesion molecules can be divided into at
least three families: the Ig gene superfamily, the selectins,
and the integrin family.s The integrin family, which is
important in dynamic regulation of adhesion and migration,
consists of a series of related ap heterodimers involved in a
variety of cell to matrix (pl and p3) and cell to cell (p2)
adhesion functions. Three subfamilies of integrins can be
distinguished by their p subunits: these are known as the
From the Division of Tumor Immunology, Dana-Farber Cancer
Institute, and the Department of Medicine, Harvard Medical School,
Boston, MA.
Submitted Feburary 6, 1992; accepted July 7, 1992.
Supported by National Institutes of Health Grant No. CA.50947.
Address reprint requests to Kenneth C. Anderson, MD, Division of
Tumor Immunology, Dana-Farber Cancer Institute, 44 Binney St,
Boston, M A 02115.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section I734 solely to
indicate this fact.
0 1992 by The American Society of Hematology.
0006-4971I92 18009-0022$3.00/0
2306
Blood, Vol80, No 9 (November I), 1992:pp 2306-2314
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
ADHESION MOLECULES ON MYELOMA CELLS
2307
Table 1. lntegrin Family of Adhesion Molecules
Subunits
Names
Ligands
a1 p l
a2pl
a3pl
a4pl
a561
a6pl
aLP2
aMp2
CD-lCD29, VLA-1
CD49blCD29, VLA-2
CD-lCD29, VLA-3
CD49dlCD29, VLA-4
CD-lCD29, VLA-5
CD49flCD29, VLA-6
CDllalCD18, LFA-1
CDllblCD18. Mac-1, CR3
CD1 lclCD18, p150.95
CD41ICD61, gpllbllla
CD51lCD61, VNR
LM, CO
LM, CO
FN, LM, CO
FN, VCAM-1
FN
LM
ICAM-1.2
CAM-1, C3bi. FX, FB
undefined
FB, FN, vWF
VN, FB, vWF, TS, FN
UP2
allbp3
aVp3
Abbreviations: FX,factor X; F6,fibrinogen; LM, laminin; CO, collagen;
FN. fibronectin; vWF, von Willebrand's factor; VN, vitronectin; TS,
thrombospondin.
their phenotypic expression of adhesion molecules. Moreover, adhesion to fibronectin was inhibited by treatment
with anti-pl integrin monoclonal antibody (MoAb) or
anti-a4 integrin MoAb, and further blocked by coculture
with RGD peptide, suggesting that these cell lines adhered
specifically to fibronectin through at least two mechanisms,
VLA-4 and RGD peptide. Finally, interleukin-6 (IL-6)
treatment of IM-9 cells decreased cell surface expression of
p l and a 4 integrins, with concomitant reduction in binding
to fibronectin. These studies suggest that myeloma cell lines
bear receptors for fibronectin and lack receptors for collagen and laminin, localizing them in marrow and inhibiting
their transit of endothelial basement membranes and exit to
the peripheral blood. They provide a model for further
study of the localization and trafficking of myeloma cells
within the marrow microenvironment.
MATERIALS AND METHODS
The U266," RPMI-8226,18 HS Sultan,19 IM-9," and ARH-77z1
human myeloma-derived cell lines and the SW626 human ovarian
cancer cell linez2were obtained from the American Type Culture
Collection (Rockville, MD). Cells were cultured in RPMI 1640
containing 10% to 15% fetal bovine serum (FBS), 100 U/mL
penicillin, and 100 pg/mL streptomycin (GIBCO, Grand Island,
NY).
Phenotypic analysis of myeloma cell lines. The cell surface
expression of adhesion molecules was determined using specific
MoAbs (Table 2)23-37in indirect immunofluorescence assays as
previously described.38 Anti-pl integrin (4B4) MoAb and anti-a4
integrin (8F2) MoAb were kindly provided by Chikao Morimoto
(Dana-Farber Cancer Institute). Anti-d integrin (12F1) MoAb
was obtained from Virgil Woods (University of California, San
Diego). Anti-a3 integrin (5143) MoAb was a gift from Anthony
Albino (Memorial Sloan Kettering Cancer Center, New York,
NY).Anti-a4 integrin (HP2/1) MoAb was obtained from AMAC,
Inc (Westbrook, ME). Anti-aS integrin18 MoAb was provided by
Kenneth Yamada (National Institutes of Health, Bethesda, MD),
and anti-a6 integrin (GoH3) MoAb was kindly provided by Arnold
Sonnenberg (Red Cross Transfusion Services, Amsterdam, Netherlands). Anti-p2 integrin (CD18) MoAb, anti-aL integrin (2F12)
MoAb, and anti-CD56 ("1)
MoAb were obtained from Jerome
Ritz (Dana-Farber Cancer Institute). Anti-aM integrin (Mo-1)
MoAb was obtained from Coulter Immunology (Hialeah, FL), and
anti-aX integrin (Leu M5) MoAb obtained from Becton Dickinson
Monoclonal Center, Inc (Mountain View, CA). Anti-CD44 (515)
MoAb was a gift from Geoffrey Kansas (Dana-Farber Cancer
Institute). A n t i - 4 3 (LM609) MoAb was kindly provided by
David Cheresh (Scripps Clinic, La Jolla, CA). Anti-aIIbp3 integrin
(P2) MoAb was obtained from AMAC, Inc. Antibody-coated cells
were enumerated by flow cytometric analysis using an EPICS V cell
sorter (Coulter Electronics, Hialeah, FL).
Assays of adhesion to ECMs. Human plasma fibronectin (FN),
human type I collagen (HC), bovine type I collagen (BC), and
mouse EHS laminin (L) were obtained from Collaborative Research, Inc (Lexington, MA). ECMs were suspended at 0.04
mg/mL in 0.01 mol/L acetic acid (HC and BC) or RPMI-1640
media (FN and L). Ninety-six-well microtiter plates (Nunc, Denmark) were coated with 20 p L of ECMs, air dried, and washed with
1% bovine serum albumin (BSA, Sigma Chemicals, St Louis, MO)
in RPMI-1640 media. Nonspecific binding was blocked by incubation with 10 mg/mL heat denatured BSA in phosphate-buffered
saline (PBS) for 1 hour at 37°C. Control wells were coated with
heat denatured BSA only.
Myeloma cell lines were radiolabeled with W r (1 mCi/mL) in
saline for 1 hour. Cells were suspended at 5 x lo5 cells/mL in
RPMI-1% BSA, and 100 p L was added to each well. After
incubation at 37°C for 30 minutes, the nonadherent cells were
removed by aspiration and inverted centrifugation (1,500 rpm for
10 minutes) or washing with 1% BSA in RPMI media. Remaining
bound cells were lysed with 0.1% NP-40 in distilled water, and the
radioactivity in cell lysates analyzed using a gamma counter (LKB,
WallacOy, Finland).
To characterize the binding of myeloma cell lines to ECMs,
antibodies known to specifically abrogate binding were used. Each
cell line was preincubated and cultured with 4B4 MoAb (1:50),
HP2/1 MoAb (20 pg/mL), LM 609 (1:1,000), P2 (l:lO), GRGDSP
peptide (1 mg/mL), or GRGESP peptide (1 mg/mL) in an attempt
to block binding to fibronectin. GRGDSP and GRGESP peptides
were obtained from Telios Pharmaceuticals, Inc (San Diego, CA).
Myeloma cell lines were next cultured with mitogens or growth
factors to characterize their effects on adherence of cell lines to
ECMs. The effect of IL-6 on adhesion was determined because
IL-6 is known to be a growth factor for myeloma cell
as
well as myeloma ~ e l l s . ~Recombinant
~ , ~ ~ - ~ IL-6 was kindly provided
by Genetics Institute (Cambridge, MA). Each cell line was preincubated and cultured with IL-6 (20 U/mL) in the ECM adherence
assay. Cell lines were also cultured with 15 ng/mL phorbol ester
Table 2. Identificationof Cell Surface Adhesion Molecules
Adhesion Molecule
MoAb
Reference
p l integrin, CD29
p2 integrin, CD18
a2 integrin, CDw49b
a3 integrin, CDw49c
a4 integrin, CDw49d
a4 integrin, CDw49d
a5 integrin, CDw49e
a6 integrin, CDw49f
UL integrin, CD1 l a
a M integrin, CD1 l b
aX integrin, CD1 IC
avp3 integrin, CD51
allbp3 integrin (CD41)
CD56
CD44
484
10F12
12F1
J143
8F2
HP211
16
GOH3
2F12
Mo-1
LeuM5
LM609
P2
NKH-1
515
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Monoclonal antibodies were used in an indirect immunofluorescence
assay to identify cell surface expression of adhesion molecules. They
were also used in blocking experiments to determine specificity of cell
line binding to ECM proteins.
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
UCHIYAMA ET AL
2308
myristate acetate (PMA), obtained from Sigma Chemicals, because PMA has been reported to modify cellular adhesion mediated by a5pl and a4pl integrin~.~’
Specific binding to ECMs was
assayed as described above. In addition, cell lines were examined
for the percentage and intensity of pl, a3, a4, and 015 integrin
expression before and after culture with IL-6 or PMA.
RESULTS
Expression of adhesion molecules on human myeloma cell
lines. Five human myeloma cell lines were examined for
their cell surface expression of adhesion molecules, including p l and p2 integrins as well as CD56 and CD44, using
indirect immunofluorescence techniques (Table 3). The
majority (84% to 98%) of cells of the U266, ARH-77, IM-9,
and HS-Sultan lines strongly reacted with MoAbs directed
at p l integrin and a4 integrin, confirming the presence of
VLA-4 on their cell surface. RPMI-8226, in contrast,
expressed lower levels of p l (41%) and a4 (13%) integrins,
suggesting that p l integrin may be associated with another
a subunit on these cells. Indeed, a 3 integrin is expressed on
10% of RPMI-8226 cells as well as a minority of U266
(22%) and IM-9 (15%) cells. None of the cell lines express
1x2, a5,or a 6 integrins. The majority ( 2 85%) of ARH-77,
IM-9, and HS-Sultan cells reacted with MoAbs directed at
p2 and a L integrins, whereas only 25% of RPMI-8226 cells
and few, if any, U266 cells do so. a M integrin is expressed
on 55% HS-Sultan cells, but on fewer than 20% cells of the
other lines examined. One third of HS Sultan cells bear ax
integrins, but the other cell lines lack this adhesion molecule. Significant expression of vitronectin receptor avp3
(CD51) and aIIbp3 (CD4la), known to be receptors for the
ligand fibronectin, was noted only on IM-9 cells (22% avp3
positive) and ARH-77 cells (67% aIIbp3 positive). CD44,
an adhesion molecule that mediates hyaluronate or collagen binding, was expressed on the majority (> 95%) of
U266, ARH-77, and IM-9 cells and on nearly one half of HS
Sultan cells. Finally, HS Sultan was the only cell line to
significantly (37% cells) express CD56, the NCAM moleTable 3. Cell Surface Expression of Adhesion Molecules on Human
Myeloma Cell Lines
% of Cells Expressing Antigen*
Antigen
U226
p1 integrin (CD29)
a2 integrin (CDw49b)
a3 integrin (CDw49c)
a4 integrin (CDw49d)
a5 integrin (CDw49e)
a6 integrin (CDw49f)
p2 integrin (CD18)
aL integrin (CDI l a )
aM integrin (CDI I b )
aX integrin (CDI IC)
avp3 integrin (CD51)
allbp3 integrin (CD4la)
CD44
CD56
97
0
22
98
0
1
6
4
3
0
2
8
98
7
RPMI-8226
ARH-77
IM9
HS Sultan
41
0
10
13
0
1
25
24
15
0
0
3
20
24
84
0
1
97
0
4
87
85
9
1
93
0
15
97
2
1
98
96
20
11
22
11
94
19
94
0
0
92
0
1
96
98
55
37
0
3
43
37
7
67
97
13
*Percentage of cells expressing antigen was quantitated by indirect
immunofluorescence flow cytometry.
cule that binds to other NCAM bearing cells, thereby
mediating homotypic adhesion.
Characterization of adhesion of myeloma cell lines to
extracellular matrix proteins. The myeloma cell lines were
assayed for their adhesion to ECMs including fibronectin,
laminin, as well as human and bovine collagen. As can be
seen in Fig 1, the U266, ARH-77, HS-Sultan, and IM-9 cell
lines adhered to fibronectin (25% to 42% specific binding),
whereas RPMI-8226 cells did not bind to this ECM.
Minimal (5% to 10%) binding of U266 and IM-9 cells was
noted to bovine collagen and laminin. Otherwise no binding
of the cell lines to human collagen, bovine collagen, or
laminin was observed. The SW626 human ovarian cancer
cell line adhered to human collagen, bovine collagen, and
laminin (28% to 49% specific binding), but did not bind to
fibronectin and, therefore, served as a control for specific
ECM binding.
To confirm that the binding of myeloma cell lines to
fibronectin was mediated by p l integrin, attempts were
made to block adherence to fibronectin by both preincubation and culture of cell lines with the anti-4B4 (CD29)
MoAb, which neutralizes p l integrin related binding. The
adhesion of U266, ARH-77, IM-9, and HS-Sultan cells to
fibronectin was inhibited approximately 50% by anti-4B4
MoAb, confirming a role for the p l integrins in this process
(Fig 2).
To further define the molecules involved in fibronectin
binding, U266 cells were preincubated and cultured with
anti p l integrin (4B4) MoAb, apti-a4 integrin (HP 2/1)
MoAb, and/or RGD peptide at titers known to maximally
block adhesion. As can be seen in Fig 3, specific binding of
U266 cells to fibronectin in the presence of anti-pl (4B4)
MoAb, anti-a4 (HP 2/1) MoAb, and excess RGD peptide
was 75%, 75%, and 50% blocked, respectively. Excess RGE
peptide, which differs from RGD peptide by a single amino
acid and does not serve as a binding site on fibronectin, was
used as a control and did not abrogate binding of U266 cells
to fibronectin. Anti-avp3 (LM609) MoAb and anti-aIIbp3
(P2) MoAb, which bind to RGD peptide on fibronectin,
also had no effect on binding of U266 to fibronectin. The
combination of 4B4 MoAb plus RGD peptide inhibited
binding to fibronectin by 80%, a significant increase in
blocking ( P = .015) relative to that noted with RGD peptide alone (Fig 3). Inhibition of binding was similarly
increased to 95% by culture with HP 2/1 MoAb and RGD
peptide, also a significant increase in blocking (P = . O S )
relative to that noted with RGD peptide alone. Conversely,
culture of cells with either HP 2/1 MoAb plus RGD
peptide or with 4B4 MoAb plus RGD peptide did not
significantlyaugment inhibition of fibronectin binding relative to that noted in the presence of 4B4 MoAb or HP 2/1
MoAb alone. Finally, preincubation and culture of U266
cells with either anti-avp3 (LM 609) or aIIbp3 (P2) MoAbs
did not abrogate binding to fibronectin.
The effect of preincubation and culture of cell lines with
IL-6, known to trigger DNA synthesis by these cells,4l or
with PMA, shown to alter adhesion mediated by d p l and
a4pl integrin~,~’
on adhesion to fibronectin was next
examined. As can be seen in Fig 4, treatment with IL-6
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
2309
ADHESION MOLECULES ON MYELOMA CELLS
T
I
uL;
Yc
v,
c-0
human collagen bovine collagen
laminin
libroneclin
EXTRACELLULAR MATRIX PROTEINS
Fig 1. Adhesion of human myeloma-derlved cell lines to ECMs.
Each Whromlum-labeled cell line was added to extracellular matrix
protein-coated wells and incubated for 30 minutes at 30°C. After
removingnonadherent cells, radioactivity in bound cells was counted
with a y counter. Percentage of specific binding (mean 2 SD from
triplicate samples) from one of four experiments is shown. The five
U266, (7)ARH-77, (0)
HSmyeloma cell lines studied include: 1.
Sultan, (n) IM-9, and (L:)RPMI-8226. The SW626 human ovarian
cancer cell line (9)adhered to human collagen, bovine collagen, and
Iaminin (28% to 49% specific binding) but did not bind to fibronectin
and, therefore, served as a control for specific ECM binding.
reproducibly decreased binding of IM-9 cell lines to fibronectin relative to that noted in media alone, without
significant effect on binding of either U266, HS Sultan, or
ARH-77 cells to fibronectin. Specifically, preincubation and
culture with IL-6 decreased specific binding of IM-9 cells to
fibronectin by 52% (30% f 6% to 15% 2 6%. n = 6,
60
'O
1
T
U266
ARH.77
HS-Sullan
IM.9
MYELOMA CELL LINES
Fig 2. 131intdn-mediaed adhesion of myeloma cell lines. Each
cellline
was pmimubated and cuhred during the adhesion
or anti-p1 (484) MoAb (E)and specific binding to
with media 1.
mmDlesb. .
fibronedin measured. Dah
SD from triDlicate
shown are from one of three experiments.
CULTURE CONDITIONS
Fig 3. Mechanism of U266 binding to fibronectin. Tha U266 call
lines showed 42% specific binding to fibronectin, showed only
minimal (5% to 10%) binding to bovine collagen and laminin, and did
not bind to human collagen. Specific binding to fibronedin was
normalized to 100% to accurately show the percentage of Inhibition
of binding to fibronedin under various culture conditions and to
thereby elucidate the mechanism of binding. The U266 cell line was
preincubated and cultured with anti p1 (484) MoAb, anti a4 (HPZ/l)
MoAb, RGD peptide, RGE peptide, 484 MoAb plus RGD peptide,
HP2/1 MoAb plus RGD peptide, anti-433 (LMBOO) MoAb and antiailblp3 (P2) MoAb. Percentage of specific binding of U266 cells to
fibronectin was measured. Data (mean 2 SD of trlplicate samples)
shown are from one of three experiments.
P = .MI). Although pretreatment and culture of ARH-77
cells with IL-6 also decreased specific binding to fibronectin
by 41% (32% +- 5% to 19% 2 15%), this was not significantly different (P= .228). Culture of cell lines with PMA
resulted in increased adhesion to fibronectin, with increments of 8% for IM-9 cells (n = 6, P = .076), of 15% for
HS-Sultan cells (n = 3, P = .047), of 19% for U266 cells
(n = 3, P = .Ml), and of 34% for ARH-77 cells (n = 3,
P = .M2) (Fig 4).
The effect of similar treatment with IL-6 or PMA on PI,
a3,a4,and a5 integrin cell surface expression on these cell
lines was measured using indirect immunofluorescenceflow
cytomctry (Table 4). Culturc of these cell lines with IL-6 or
PMA did not alter ccll surfacc expression of a3 or a5
integrins. Moreover, treatment with IL-6 lead to only a 13%
2 6% (n = 3, P = .292) and 13% 2 2% (n = 3,P = .031)
decrease in p1 and a4 integrins, respectively. Although
decreases in pl and a4 integrin expression on ARH cells
were observed after IL-6 treatment (Table 4),these changes
were neither reproducible nor significant. There were no
effects of IL-6 treatment on cell surface p1 or a4 integrin
exmession on U266. RPMI-8226. and HS-Sultan cells.
b e decrease in binding of IM-9 and ARH-77 cell to
fibronectin after IL-6 treatment was also accompanied by
changes in intensity of integrin expression on these cells. As
can hc seen in Fig 5A, thc mean peak fluorescence intensity
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
UCHIYAMA ET AL
2310
T
T
U266
AR H. 7 7
HS.Sullan
IM-9
MYELOMA CELL LINE
Fig 4. Effects of lL-6 and PMA on cell line adhesion. Each cell line
was preincubated and cultured with media )1.
or with interleukin-6
(20 U/mL) (a)and added directly to the extracellular matrix protein
coated wells. PMA (15 ng/mL) (0) was added to the wells concomitantly with cells. Percentage of specific binding of cells to fibronectin
was measured. Data (mean 2 SD of triplicate samples) shown are
from one of three experiments.
of p1 and a4 integrin on IM-9 was dccrcascd by treatmcnt
with IL-6. from 5.1 to 2.9 for p1 intcgrin and from 4.9 to 2.9
for a4 intcgrin. Similarly, the mcan peak fluoresccncc
intensity of PI and a4 intcgrin exprcssion on ARH-77 cells
decreased aftcr stimulation with IL-6from 4.0 to 3.8 for PI
intcgrin and 5.2 to 4.2 for a4 intcgrin (Fig SB). In contrast,
IL-6 trcatmcnt did not alter thc intcnsity of p1 or a4
intcgrin exprcssion on U266 and HS Sultan cells. Culture of
ccll lines with PMA did not altcr thc pcrccntagc of cells
bearing pl or a4 intcgrins on any mycloma ccll lincs (Tablc
4). Howcvcr, a slight dccreasc in intcnsity of both Pl and a4
intcgrin exprcssion was observed on IM-9 cells aftcr PMA
trcatmcnt (Fig SA).
DISCUSSION
Adhesion molecules play an important role in migration
and homing of mononuclcar cells to facilitate cell to ccll
intcractions in the normal immune
In hcmatologic malignancics, adhcsion molecules may also bc crucial
in determining thc propcrtics rcquircd for rclcasc of
leukemic cclls from thc BM. Indccd thc scqucncc of
adhcsion and dctachmcnt is fundamcntal to ccll locomotion and csscntial to malignant invasion.4x In multiplc
mycloma, the malignant plasma cells arc localized in thc
BM and rarcly circulate in thc pcriphcral blood. Howevcr,
little is known of thc mechanisms by which malignant
plasma cells adhere to thc BM stroma and how these
usually scssile cells uncxpectcdly appear in thc blood.
Although normal plasma cclls adhcrc tightly to thc marrow
stroma and do not tcnd to migratc, malignant plasma cclls
can bc phenotypically divcrgcnt, ic, subscts arc ablc to
dctach from thc bonc marrow stroma and migratc to
cxtramcdullary sitcs in plasma ccll leukemia? Thus, it is
important to charactcrizc thc adhesive propcrties of malignant cells that arc localizcd in BM and to contrast thcm
with malignant cclls that have a propensity to invadc
basement mcmbrancs and thereby exit from marrow.
Normal pcriphcral blood B cells bcar VLA-4; howcvcr,
thc lcvcl of a4 subunit is almost twice the lcvcl of PI
dctcctcd, suggcsting that not all of the a4 subunit is
associated with PI." Thc cxprcssion of p1 intcgrins on
normal plasma cclls is not yct dcfincd. Our studies showed
that four human mycloma-dcrivcd ccll lincs also cxprcsscd
VLA-4 protcin; howcvcr, in contrast to normal B cclls, the
lcvcls of a4 and pl intcgrins on thcsc cclls is cquivalcnt.
Thcsc cclls cxprcssing VLA-4 protcin adhcrcd to fibroncctin, and this adhcrcncc could bc partially blockcd with
cithcr anti-pl intcgrin or anti-a4 intcgrin MoAbs. Morcovcr, thc binding of thcsc cclls to fibroncctin was also
partially blockcd with RGD pcptidc, suggcsting that their
binding to fibroncctin is mcdiatcd not only by VLA-4
protcin but also by othcr receptors with fibroncctin binding
propcrtics, such as vitroncctin rcccptor (aV intcgrin) or gp
IIbIIIa. Phcnotypic and functional studics confirm that
avp3 and aI1bp.t are not involved in binding of myeloma
ccll lincs to fibroncctin, but othcr vitroncctin receptors, ic,
avpl or avpS, remain as possible mediators of fibroncctin
binding. Importantly, specific binding to fibronectin could
bc nearly complctcly blockcd by the addition of anti-pl or
anti-a4 intcgrin MoAb to cxccss RGD pcptidc. Thcsc data
suggest that thc ability of mycloma cclls to bind to fibroncctin may contributc to their localization in the bone marrow
and, convcrscly, that the loss of fibroncctin receptors may
correlate with a propensity to cmigratc from thc bonc
marrow microcnvironmcnt. Finally, thcsc cclls do not
Table 4. Effect of PhorbolEster Mytistate Acetate or Interleukin-8
Treatment on Cell Surface Expression of Adhesion Molecules on
Myeloma Cell Lines
Cell Line
u266
IM-9
ARH-77
RPMI-8226
HS-Sultan
Culture
Conditions
Media
PMAt
IL-6*
Media
PMA
IL-6
Media
PMA
IL-6
Media
PMA
IL-6
Media
PMA
IL-6
% of Cells Expressing Antigen'
pl lnlegrin a3 lntegrin
99
99
96
94
a8
78
70
89
48
19
14
17
83
74
74
17
15
26
1
0
2
0
0
0
4
3
0
1
0
0
w4 lnlegrin
99
98
96
92
85
78
87
94
58
3
4
0
82
72
72
a5 lnlegrin
0
0
0
0
0
3
0
0
0
4
10
0
1
0
0
Cells were cultured with IL-6 or with PMA.
'Percentage of cells expressing antigen was quantitated by indirect
immunofluorescenceflow cytometry.
tPhorboi ester myristate acetate.
Slnterleukin-6.
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
2311
ADHESION MOLECULES ON MYELOMA CELLS
P
CONTROL
-
PMATREATMENT
.
IL-6
TREATMENT
..
r:
k
D
m
P
a
I
E
91
/'
z
d
J
W
0
E
k
tr
e
P
E
rl
t
U
10%
lo*
LOGFLUORESCENCE I N T E N S I T Y
IPI-9
E
CONTROL
-
PI*LA TREAT-T
SL-6
.__
TREATMENT
-
E
k
b
II
Y
Fig 5. Alterations in intensity
of pl and a4 integrin expression.
(A) The IM-9 and (B) ARH-77 cell
lines were preincubated and cultured with media, IL-6 (20 U/mL)
or PMA (15 mg/mL). Percentage
of cells expressing 61 and OA
integrins was determined using
the 484 and HP 21 1 MoAbs, respectively, in indirectimmunofluorescence assays. Antibody
coated cells were enumerated by
flow cytometric analysis using
an EPICS V cell sorter (Coulter
Electronics). Histogram displays
cell number (y axis) versus log
fluorescence intensity(x axis).
a
I
C
rl
0
2
cl
C
w
k
J
m
U
B
Y
E
t
I
10'
102
adhere to collagen and laminin. Because the basal lamina is
composed mainly of type IV collagen and lamini11,4~the
lack of receptors for these ECM components also prohibits
their transit to extravascular spaces.
The high serum levels of IL-6 noted in patients with
active
coupled with the recent observation that
systemic administration of anti-IL-6 may inhibit growth of
myeloma cells in vivo,44strongly supports a role for IL-6 in
the pathophysiology of myeloma. Previous studies also
suggest a role for IL-6 in growth of human multiple
myeloma-derived cell lines."41 In the present study, we
showed that IL-6 treatment of the IM-9 myeloma cell lines
decreased binding to fibronectin, with a concomitant decrease in VLA-4 cell surface expression. Although the
influence of ECM proteins and integrins on the behavior
and development of normal plasma cells has not yet been
108
10,
108
101
LOGFLUORESCENCE I N T E N S I T Y
ARH- 7 7
studied, it is known that normal pre-B cells appear to use
two different integrin-like fibronectin receptors during
migration and maturation in the BM, before losing this
capacity upon release into
Similarly, maturation of
reticulocytes within the BM is associated with loss of
adhesion to fibronectin as well as a loss of detectable
VLA-Mike complex>l In these examples, it is clear that
matrix-to-integrin interactions play a significant role in the
development of blood cells. We hypothesize that some
malignant plasma cells remain in the BM because of
binding to fibronectin. Moreover, changes in adhesion of
myeloma cells which are induced by IL-6 in vitro may reflect
the in vivo effects of IL-6 on malignant plasma cells, ie, their
ability to exit from the BM and circulate in plasma cell
leukemia. In our studies, IL-6 induced only small changes in
p l and a4 integrin expression. Decreased adhesion to
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
UCHIYAMA ET AL
2312
fibronectin induced by IL-6 may, therefore, be related to a
decrease in affinity of fibronectin receptors.
In the present study, we examined the effect of PMA on
expression and function of p l integrins on myeloma cell
lines. Treatment with PMA enhanced binding to fibronectin on all myeloma cell lines, but did not upregulate either
p l or 014 integrin expression. Therefore, the effect of PMA
on cell surface expression was dissociated from its effects on
receptor function, and PMA appears to induce an increased affinity for binding to fibronectin. The mechanism
whereby PMA enhances integrin-mediated enhancement
of cellular adhesion has not been defined. Changes in the
functional state of integrins can be induced by PMA, ie,
phorbol esters dramatically upregulate binding mediated by
01Lf32 as well as aMP2 integrin.5,s2Activation with PMA
may also result in conformational changes of integrin within
the cytoskeleton, thereby altering the distribution of receptorss3:for example, treatment with PMA led to localization
of talin with p l integrin in antibody-induced caps on the
surface of lymphocytes, consistent with a metabolically
regulated change in membrane-cyto~keleton.~~
It has been
reported that VLA-4 on B cells mediates binding to
germinal center cells only when B cells are activated, ie,
with PMA.55 The effects of PMA treatment on adherence of
malignant cells are not as well studied. Wilkins et aIs6
reported that the enhanced adherence of human T-cell
leukemia cells to ECMs, induced by PMA, did not involve
upregulation of receptor expression, but rather was derived
from the increased affinity of binding for structures that
were expressed on the cell surface. In other studies of p l
integrins, PMA has induced a loss of VLA-5 glycosylation,a
loss of VLA-5 function, or an increase in VLA-5 expression
on leukemic cell lines>7 PMA has also been shown to
induce differentiation as well as enhanced adhesion to
fibronectin, associated with modulation of VLA-4 and
VLAJ fibronectin re~eptors.4~
Finally, recent studies suggest that follicular non-Hodgkin's lymphomas adhere to
normal germinal centers via VLA-4 on the tumor cell
surface.58Our studies suggest that VLA-4 may also mediate
binding of myeloma cell lines to fibronectin. Moreover, they
further suggest that VLA-4 can be upregulated functionally
by PMA in vitro, suggesting that they may also be capable of
similar functional upregulation in vivo.
The present studies used myeloma-derived cell lines,
which may not reflect either the cell surface phenotype or
biology of myeloma in vivo.16Therefore, in future studies it
will be necessary to characterize the function of adhesion
molecules on freshly isolated tumor cells from patients with
multiple myeloma. Such studies will define mechanisms of
adherence not only to ECMs but also to myeloma stromal
cells and, thereby, will enhance our understanding of the
localization and trafficking of cells within the BM microenvironment.
ACKNOWLEDGMENT
We thank Bernadette Miner for assistance in the preparation of
this manuscript.
REFERENCES
1. Mellstedt H, Holm G, Pettersson D, Peest D: Idiotypebearing lymphoid cells in plasma cell neoplasia. Clin Haematol
11:65,1982
2. Pilarski LM, Mant MJ, Reuther B A Pre-B cells in the
peripheral blood of multiple myeloma patients. Blood 66:416,1985
3. Jensen GS, Mant MJ, Belch AJ, Berenson JR,Ruether BA,
Pilarski LM: Selective expression of CD45 isoforms defines CALLA+ monoclonal B lineage cells in peripheral blood from myeloma
patients as late stage B cells. Blood 78:711,1991
4. Kalasz V, Pap T, Csanaky G, Kelenyi G: Endothelial cell
receptors on leukemic plasma cells. Leuk Res 13:863,1989
5. Springer TA: Adhesion receptors of the immune system.
Nature 346:425,1990
6. Hemler ME: Integrin family: Structures, functions, and their
role on leukocytes. Annu Rev Immunol8:365,1990
7. Pilarski LM, Jensen GS, Mant MJ, Belch AR: Abnormal
monoclonal B lineage cells in blood may comprise the malignant
stem cell in myeloma: Expression of pl and p2 integrins. Proc 111
International Workshop on Multiple Myeloma. Torino, Italy, 1991,
P 58
8. Marlin SD, Springer TA: Purified intercellular adhesion
molecule (ICAM-1) is a ligand for lymphocyte function-associated
antigen (LFA-1). Cell 51:813,1987
9. Staunton DE, Dustin ML, Springer TA Functional cloning of
ICAM-2, a cell-adhesion ligand for LFA-1 homologous to ICAM-1.
Nature 33958,1989
10. Springer TA, Dustin ML, Kishimoto TK, Marlin SD: The
lymphocytefunction-associated LFA-1, CD2, and LFA-3 molecules:
Cell adhesion receptors of the immune system. Annu Rev Immunol
5:223,1987
11. van Noesel C, Miedema F, Brouwer M, de Rie MA, Aarden
LA, van Lier RAW Regulatory properties of LFA-1 alpha and
beta chains in human T lymphocyte activation. Nature 333:850,
1988
12. Boyd AW, Wawryk SO, Bums GF, Fecondo Jv: Intercellular adhesion molecule 1 (ICAM-1) has a central role in cell-cell
contact-mediated immune mechanisms. Proc Natl Acad Sci USA
85:3095, 1988
13. Dustin ML, Springer T A Lymphocyte function-associated
antigen-1 (LFA-1) interaction with intercellular adhesion molecule 1 (ICAM-1) is one of at least three mechanisms for lymphocyte adhesion to cultured endothelial cells. J Cell Biol 107:321,
1988
14. Hamann A, Jablonsky-Westrich D, Duijvestijn A, Butcher
EC, Baisch H, Harder R, Thiele H: Evidence for an accessory role
of LFA-1 in lymphocyte-high endothelium interaction during
homing. J Immunol140:693,1988
15. Ahsmann EJM, Lokhorst HM, Dekker AW, Bast BJEG,
Bloem AC: Lymphocyte function-associated antigen-one (LFA-1)
expression on plasma cells correlates with tumour growth in
multiple myeloma. Blood 79:2068,1992
16. Hamilton MS, Ball J, Bromidge E, Franklin IM: Specific
antigen expression of human neoplastic plasma cells includes
molecules associated with lymphocyte recirculation and adhesion.
Br J Haematol78:60,1991
17. Nilsson K, Bennich H, Johansson S, Ponten J: Established
immunoglobulin producing myeloma (IgE) and lymphoblastic
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
ADHESION MOLECULES ON MYELOMA CELLS
(IgG) cell lines from an IgE myeloma patient. Clin Exp Immunol
7:477, 1970
18. Matsuoka Y, Moore G, Yagi Y, Pressman D: Production of
free light chains of immunoglobulin by a hematopoietic cell line
derived from a patient with multiple myeloma. Proc SOCExp Biol
Med 125:1246,1967
19. Harris N: Plasma cell surface antigen on human blood
lymphocytes. Nature 250507,1974
20. Fahey J, Buell D, Sox H: Proliferation and differentiation of
lymphoid cells: Studies with human lymphoid cell lines and
immunoglobulin synthesis. Ann N Y Acad Sci 190:221, 1971
21. Burk K, Drewinko B, Trujillo J, Ahearn M: Establishment of
a human plasma cell line in vitro. Cancer Res 38:2508,1978
22. Fogh J, Fogh JM, Orfeo T One hundred and twenty-seven
cultured human tumor cell lines producing tumors in nude mice. J
Natl Cancer Inst 59:221,1977
23. Morimoto C, Letvin NL, Boyd AW, Hagan M, Brown HM,
Kornacki MM, Schlossman S F The isolation and characterization
of the human helper inducer T cell subset. J Immunol 1343762,
1985
24. Robertson MJ, Caligiuri MA, Manley TJ, Levine H, Ritz J:
Human natural killer cell adhesion molecules. J Immunol 145:
3194,1990
25. Pischel KD, Hemler ME, Huang C, Bluestein HG, Woods
V L Use of the monoclonal antibody 12F1 to characterize the
differentiation antigen VLA-2. J Immunol 138:226,1987
26. Fradet Y, Cordon-Cardo C, Thomson T, Daly ME, Whitmore WF Jr, Lloyd KO, Melamed MR, Old U:Cell surface
antigens of human bladder cancer defined by mouse monoclonal
antibodies. Proc Natl Acad Sci USA 81:224,1984
27. Nojima Y, Humphries MJ, Mould AP,Komoriya A, Yamada
KM, Schlossman SF, Morimoto C: VLA-4 mediates CD3dependent CD4+ T cell activation via the CS1 alternatively spliced
domain of fibronectin. J Exp Med 172:1185,1990
28. Sanchez-Madrid F, De Landazuri MO, Morago G, Cebrian
M, Acevedo A, Bernabeu C: VLA-3: A novel polypeptide association within the VLA molecular complex: Cell distribution and
biochemical characterization. Eur J Immunol16:1343, 1986
29. Matsuyama T, Yamada A, Kay J, Yamada KM, Akiyama SK,
Schlossman SF, Morimoto C: Activation of CD4 cells by fibronectin and anti-CD3 antibody. J Exp Med 1701133,1989
30. Sonnenberg A, Janssen H, Hogervorst F, Calafat J, Hilgers
J: A complex of platelet glycoproteins ICand IIa identified by a rat
monoclonal antibody. J Biol Chem 26210376,1987
31. Tedder TF, Schmidt RE, Rudd CE, Kornacki MM, Ritz J,
Schlossman S F Function of the LFA-1 and T4 molecules in the
direct activation of resting human B lymphocytes by T lymphocytes.
Eur J Immunol16:1539,1986
32. Todd R F 111, Nadler LM, Schlossman SF: Antigens of
human monocytes identified by monoclonal antibodies. J Immunol
126:1435,1981
33. Lanier LL, Arnaout MA, Schwarting R, Warner NL, Ross
GD: p150/95, third member of the LFA-l/CR3 polypeptide family
identified by anti-Leu M5 monoclonal antibody. Eur J Immunol
15:713, 1985
34. Cheresh DA, Smith JW, Cooper HM, Quaranta V: A novel
vitronectin receptor integrin (avPx)is responsible for distinct
adhesive properties of carcinoma cells. Cell 57:59, 1989
35. McGregor JL, Brochier J, Wild F, Follea G, Trzeciak MC,
James E, Dechavanne M, McGregor L, Clemetson K Monoclonal
antibodies against platelet membrane glycoproteins. Eur J Biochem 13:427,1983
36. Hercend T, Griffin JD, Bensussan A, Schmidt RE, Edson
MA, Brennan A, Murray C, Daley JF, Schlossman SF, Ritz J:
2313
Generation of monoclonal antibodies to a human NK clone:
Characterization of two NK associated antigens, NKHlA and
NKH2, expressed on subsets of large granular lymphocytes. J Clin
Invest 75:932,1985
37. Kansas GS, Wood GS, Dailey MO: A family of cell-surface
glycoproteins defined by a putative anti-endothelial cell receptor
antibody in man. J Immunol142:3050,1989
38. Anderson K, Jones R, Morimoto C, Leavitt P, Barut B:
Response of purified myeloma cells to hematopoietic growth
factors. Blood 73:1915,1989
39. Schwab G, Siegall C, Aarden L, Neckers L, Nordan R:
Characterization of an interleukin-6-mediated autocrine growth
loop in the human multiple myeloma cell line, U266. Blood 77:587,
1991
40. Hitzler J, Martinez-Lavdez H, Bergsagel D, Minden M,
Messner H: Role of interleukin-6 in the proliferation of human
multiple myeloma cell lines OCI-MY 1 to 7 established from
patients with advanced stage of the disease. Blood 78:1996,1991
41. Barut BA, Zon LI, Cochran MK, Paul SR, Chauhan D,
Fingeroth J, Anderson KC: Role of interleukin-6 in the growth of
myeloma-derived cell lines. Leuk Res (in press)
42. Kawano M, Hirano T, Matsuda T, Taga T, Horii Y ,Iwato K,
Asaoku H, Tang B, Tanabe 0,Tanaka H, Kuramoto A, Kishimoto
T Autocrine generation and requirement of BSF-2/IL-6 for
human multiple myelomas. Nature 332:83,1988
43. Bataille R, Jourdan M, Zhang XG, Klein B: Serum levels of
interleukin-6, a potent myeloma cell growth factor, as a reflection
of disease severity in plasma cell dyscrasias. J Clin Invest 84:2008,
1989
44. Klein B, Wijdenes J, Zhang XG, Jourdan M, Boiron JM,
Brochier J, Liautard J, Merlin M, Clement C, Morel-Fournier B,
Lu ZY, Mannoni P, Sany J, Bataille R: Murine anti-interleukin-6
monoclonal antibody therapy for a patient with plasma cell
leukemia. Blood 78:1198,1991
45. Klein B, Zhang X, Jourdan M, Content J, Houssiau F,
Aarden L, Piechaczyk M, Bataille R: Paracrine rather than
autocrine regulation of myeloma-cell growth and differentiation by
interleukin-6. Blood 73:517,1989
46. Nilsson K: The role of interleukin-6 (IL-6) as a growth factor
for human multiple myeloma cells. Proc 111 International Workshop on Multiple Myeloma. Torino, Italy, 1991, p 43
47. Ferreira 0-C, Valinsky JE, Sheridan K, Wayne EA, Bianco
C, Garcia-Pardo A: Phorbol-ester induced differentiation of U937
cells enhances attachment to fibronectin and distinctly modulates
the asp1 and q& fibronectin receptors. Exp Cell Res 193:20,1991
48. Nicolson G L Cell surface molecules and tumor metastasis.
Exp Cell Res 150:3,1984
49. Martin GR and Tinpl R: Laminin and other basement
membrane components. Annu Rev Cell Biol3:57,1987
50. Bernardi P, Patel VP, Lodish HF: Lymphoid precursor cells
adhere to two different sites on fibronectin. J Cell Biol 105:489,
1987
51. Patel VP, Lodish H F The fibronectin receptor on mammalian erythroid precursor cells: Characterization and developmental
regulation. J Cell Biol 102:449, 1986
52. Detmers PA, Wright SD, Olsen E, Kimball B, Cohn Z A
Aggregation of complement receptors on human neutrophils in the
absence of ligand. J Cell Biol105:1137,1987
53. Kupfer A, Singer SJ: Cell biology of cytotoxic and helper T
cell functions: Immunofluorescence microscopic studies of single
cells and cell couples. Annu Rev Immunol7:309,1989
54. Burn P, Kupfer A, Singer SJ: Dynamic membrane-cytoskeletal interactions: Specific association of integrin and talin arises in
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
2314
vivo after phorbol ester treatment of peripheral blood lymphocytes.
Proc Natl Acad Sci USA 85:497,1988
55. Freedman AS, Munro JN, Rice GE, Bevilacqua MP, Morimoto C, McIntyre BW, Rhynhart K, Pober JS, Nadler LM:
Adhesion of human B cells to germinal centers in vitro involves
VLA-4 and INCAM-110. Science 249:1030,1990
integrin56. Wilkins JA, Stupack D, Stewart S, Caixia S:
mediated lymphocyte adherence to extracellular matrix is enhanced by phorbol ester treatment. Eur J Immunol21:517,1991
UCHIYAMA ET AL
57. Symington BE, Symington FW, Rohrschneider LR: Phorbol
ester induces increased expression, altered glycosylation, and
reduced adhesion of K562 erythroleukemia cell fibronectin receptors. J Biol Chem 264:13258, 1989
58. Freedman AS, Munro JN, Morimoto C, McIntyre BW,
Rhynhart K, Lee N, Nadler LM: Follicular non-Hodgkin’s lymphoma cell adhesion to normal germinal centers and neoplastic
follicles involves very late antigen-4 and vascular cell adhesion
molecule-I. Blood 79906,1992
From www.bloodjournal.org by guest on July 31, 2017. For personal use only.
1992 80: 2306-2314
Characterization of adhesion molecules on human myeloma cell lines
H Uchiyama, BA Barut, D Chauhan, SA Cannistra and KC Anderson
Updated information and services can be found at:
http://www.bloodjournal.org/content/80/9/2306.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.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.