From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
Adhesion Molecule Expression on B-Cell Chronic Lymphocytic Leukemia
Cells: Malignant Cell Phenotypes Define Distinct Disease Subsets
By Giulio De Rossi, Daniela Zarcone, Francesca Mauro, Giannamaria Cerruti, Claudya Tenca, Antonio Puccetti,
Franco Mandelli, and Carlo E. Grossi
Expression of surface adhesion molecules of the lg superfamily (CD54 and CD58), of the integrin family (pl, &, and
chains), of the selectin family (L-selectin), and of the
lymphocyte homing receptor (CD44) was analyzed on 6cell chronic lymphocytic leukemia (6-CLL) cells from 7 4 patients. The aim of the study was the definition of phenotypically distinct disease subsets and the correlation of
adhesion molecule phenotypes with clinical parameters.
Expression of CD58 on 6-CLL cells defined more advanced
disease stages. In comparisonwith p chain-positive cases,
patients whose cells did not express B1, &, and p3 integrin
chains fell into the most favorable prognostic group, with
lower lymphocytosisand the absence of splenomegaly, diffuse bone marrow infiltration, and therapy requirement. A
novel finding was the expression of p3 chains on cells from
a minority (12 of 74) of 6-CLLcases. B3 chains were always
coexpressedwith & and p2 chains. Two-color immunofluorescence analyses of adhesion molecules such as ax& integrin (LeuM5) and L-selectin (Leu8) showed that these
markers were detectable on variable proportions of leukemic cells, thus confirming the intraclonal phenotypic
heterogeneityof 6-CLL. Differencesin the intensity of CD44
expression were also shown among the various 6-CLL
clones. Finally, no major variations were shown by comparisonof adhesion molecule phenotypesof leukemic cells
simultaneouslyobtained from blood and bone marrow, and
of CD5+ versus CD5- clones.
0 1993 by The American Society of Hematology.
M
a clear correlation between the surface antigenic phenotype
of leukemic cells and the clinical behavior has not been defined.
In the present study, we have investigated the expression
of a variety of adhesion molecules on cells from 74 B-CLL
patients. Surface molecules of the Ig superfamily (CD54/intercellular adhesion molecule- 1 [ICAM- I ] and CD58/lymphocyte function-associated antigen 3 [LFA-3]); of the PI,
p2, and p 3 integrin family; of the selectin family (L-selectin/
Leu8); and of CD44 have been analyzed by immunofluorescence and flow cytometry (see the l i t e r a t ~ r e for
~ ~ reviews
.~~
on these molecules). Expression of adhesion molecules has
allowed distinction of phenotypically diverse groups of patients and clinical differences have emerged among some of
these groups.
a3
OST LYMPHOPROLIFERATIVE disorders are
monoclonal expansions of B or T cells at different
stages of maturation that evolve as leukemia or lymphoma.’
Malignant cells proliferate within the tissue of origin and
some of them reach the circulation and spread via blood and
lymph vessels. In a further step of tumor progression, these
cells extravasate and seed into other t i ~ s u e s .Adhesive
~.~
interactions between neoplastic cells, endothelial cells, and extracellular matrix (ECM) proteins are crucial for these sequential events to O C C U ~ . ~ . ~
Expression of adhesion molecules of the Ig superfamily,
of the integrin family, of the selectin family, and of the “lymphocyte homing receptor,” CD44, has been investigated in
lymphoproliferative
However, a complete pattern of adhesion molecule expression in lymphoid malignancy
has not yet emerged, and conflicting data on the prognostic
significance of these phenotypic features have been reported. ‘’,I6
B-cell chronic lymphocytic leukemia (B-CLL) is a lymphoproliferative disorder with a high degree of biologic hete r ~ g e n e i t y . ”This
. ~ ~ lymphoid malignancy originates from a
CD5+ normal B-cell counterpart that may represent a distinct
subset of B
However, in less than 10%of the cases,
B-CLL cells do not express CD5.19-2’Other phenotypic features may be heterogeneous among B-CLL cases, such as
expression of pz integrin molecules (CD I I b/CD18 and
CD1 lc/CD18) found on myelomonocytic cells and on hairy
cell leukemia (HCL) ~ e l l s .Expression
~ ~ , ~ ~ of CD 1 1b/CD 18
has been associated with the ability of B-CLL cells to produce
interleukin- 1,23 whereas expression of CD1 lc/CD I8 appears
to define a B-CLL variant with distinctive clinical feature^.^^.^^
Further phenotypic heterogeneity of B-CLL cells is due to
maturational differences among malignant clones and to intraclonal maturation, as shown by ultrastructural and cytochemical analyses.27
Clinical heterogeneity of B-CLL is also well defined and
prognosis is often unpredi~table.’’.~~
Many patients do not
require treatment and are unlikely to die of leukemia. In
contrast, other patients present with rapidly progressive disease and have a median survival of less than 2 years. So far,
Blood, Vo181, No 10(May 15), 1993: pp2679-2687
MATERIALS AND METHODS
Patients. Seventy-four patients were evaluated in our Division
from September 1990 through December 199 I. The mean age of the
patients was 62 years, with a range between 41 and 85 years. There
was one 19-year-old female. The male to female ratio was 40 to 34.
B-CLL was diagnosed according to generally accepted criteria that
included peripheral blood lymphocytosis greater than 5,OOO/pL and
bone marrow lymphocytosis greater than 3070.’‘ Diagnosis was confirmed by immunophenotypic analysis of surface Ig light chain and
From Hematology, Hlrman Biopathology Department, University
“La Sapienza”, Rome; the National Institute for Cancer Research.
Genova; and the Department of Human Anatomy, University of Genova, Genova, Italy.
Submitted September 16, 1992; accepted December 28, 1992.
Supported by grants fiom the National Research Council (CNR)
of Italy (Target Projects: BTBS and ACRO) and ,from the Italian
Association for Cancer Research (AIRC).
Address reprint requests to Carlo E. Grossi, MD, Department of
Anatomy, University of Genova, Via De Toni, 14, 16132 Genova,
Italy.
The publication costs of this article were defiayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 V.S.C. section I734 solely to
indicate this fact.
0 I993 by The American Society of Hematology.
0006-4971/93/81I O-0026$3.00/0
2679
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2680
CD5 expression. We intentionally selected a disproportionate number
of CD5- cases ( I 7 of 74) with the aim of evaluating the expression
of adhesion molecules in this relatively rare B-CLL variant.
At the time of study, 63 of 74 patients had never been treated.
However, at a later time, 23 of the 63 patients had to start treatment
with chlorambucil (CHL) and prednisone (PDN). Eleven patients
were studied when receiving the CHL plus PDN regimen.
As of June 1992, I 1 patients are deceased, 9 of disease progression,
1 due to a second neoplasia, and 1 from myocardial infarction.
Clinical evaluation of the patients included staging according to
Rai et aP5and Binet et al,36peripheral blood lymphocytosis,evaluation
of liver, spleen, and lymph node enlargement (aided by computed
axial tomography [CAT] and echotomography), and lymphocyte
doubling time (LDT).37
The pattern of bone marrow infiltrati~n~~
was evaluated in bone
marrow biopsies of 52 of 74 patients. Biopsy was not performed in
22 patients due to advanced age, cardiovascular risk factor, or refusal
of informed consent.
Immmophenotypic analyses. Mononuclear cells were isolated
from heparinized peripheral blood samples by Ficoll-Hypaque density
gradient centrifugation. When, in addition to peripheral blood samples, bone marrow aspirates were also available (7 of 74 patients),
single-cell suspensions of bone marrow cells were centrifuged on Ficoll-Hypaque density gradients to recover mononuclear cells. Freshly
isolated mononuclear cells were used for immunophenotypic analyses.
Further immunophenotypic and biochemical studies were performed
using frozen cell aliquots stored in liquid nitrogen.
Cells were incubated with mouse monoclonal antibodies (MoAbs)
to a variety of surface molecules (as indicated in Table l), followed
by affinity-purified, fluorescein isothiocyanate (FITC)-labeled goat
antimouse Ig (Southern BiotechnologyAssociates, Birmingham, AL).
In other experiments, biotinylated MoAbs were used, followed by
FITC-streptavidin.
Two-color immunofluorescence analyses were performed using
FITC-labeled and phycoerythrin (PE)-labeled mouse MoAbs (Table
1). Controls were provided by cells incubated directly with the secondary reagent, omitting the primary reagent, and by cells incubated
with an unrelated MoAb of an isotype identical to that of the test
MoAb (anti-CD62, IgG1; anti-CD10, IgG2a; anti-CD69, IgGZb),
followed by the labeled secondary reagent (Table 1). These antibodies
were used as negative controls.
Cells were analyzed by one-color or two-color fluorescence using
a FACScan flow cytometer (Becton Dickinson, San Jose, CA) and
setting gates on the lymphocyte populations. Analysis by two-color
fluorescence was restricted to those markers that appeared equivocal
as for their attribution to B-CLL cells or to residual T cells. However,
this was seldom the case because in the majority of the patients (53
of 74) the percentage of leukemic cells was greater than 70%.
Immunoprecipitation. B-CLL cells were washed twice in cold
phosphate-buffered saline (PBS) and resuspended in PBS with IO
mmol/L glucose. Surface proteins were labeled with sulfosuccinimidyl
6-(biotinamido) hexanoate (NHS-LC-Biotin; Pierce, Rockford, IL).39
Ten million cells and 2.5 mg of biotin were used per experiment.
Labeled cells were lysed with 10 mmol/L Tris, 150 mmol/L NaCI,
and 1% NP-40, pH 7.5, containing protease inhibitors. Lysates were
centrifuged for I O minutes at 10,OOOg. The supematant was precleared
twice with 50 fiL of beads coated with antimouse IgG (Calbiochem,
San Diego, CA). Beads were sedimented by centrifugation for 2 minutes at 10,OOOg and the supernatant was used for immunoprecipitation. Ten micrograms of monoclonal anti-CD61 MoAb (Immunotech) was added to each sample and incubated overnight at 4°C.
Fifty microliters of bead-immobilized antimouse IgG was added to
each tube and the samples were rotated for I hour at 4°C. After
centrifugation for 2 minutes at lO,OOOg, beads were washed three
DE ROSS1 ET AL
Table 1. Isowpe. Specificity, and Source of MoAbs
MoAb
lsotype
Specificity
-
HLA-DR (FITC)"
Leu 4 (FITC, PE)
Leu 5b
IOL 5 4
IOT 58
IOT 2 9
4B4
IOT 18
IOT 1 6
Leu 15
Leu M5 (PE)
IOP 6 1
Leu 8 (FITC, PE)
Leu 4 4 (FITC)
IOP 41a
IOP 6 2 t
IOB 5at
IOA 6 9 t
IgG 2a
IgG 1
IgG 2a
IgG 1
IgG 2a
IgG 2a
IgG 1
IgG 1
IgG 1
IgG 1
IgG 2 b
IgG 1
IgG 2a
IgG 1
IgG 1
IgG 1
IgG 2a
IgG 2b
HLA-DR
CD3
CD2
CD54 (ICAM-I)
CD58 (LFA-3)
CD29 ( P i )
~ ~ (8,)
2 9
CDI 8 (PA
CD 1 1a (aL)
CD11 b (a,.,)
CD1 IC
(ax)
CD61 (63)
L-selectin
CD44
GPllb-llla (CD41a)
P-selectin (CD62)
CDIO
CD69
BD
BD
BD
IMM
IMM
IMM
Source
co
IMM
IMM
BD
BD
IMM
BD
BD
IMM
IMM
IMM
IMM
Abbreviations: BD, Becton Dickinson (SanJose, CA); IMM, lmmunotech
(Marseille-Luminy, France); CO, Coulter (Hialeah, FL).
Fluorochromes in parentheses refer t o directly labeled MoAb used
for two-color immunofluorescence analyses.
t Unrelated MoAb used as isotype-matched negative controls.
times with cold lysis buffer and twice with 10 mmol/L Tris, 2 mmol/
L ECTA, 0. I % sodium dodecyl sulfate (SDS), pH 7.5. followed by
elution with sample buffer. Eluates were run on 7.5% polyacrylamide
gel under nonreducing conditions. Polypeptides separated by SDSpolyacrylamide gel electrophoresis (SDS-PAGE) were transferred to
nitrocellulose. The filter was saturated for I hour with 50 mmol/L
Tris, 150 mmol/L NaCI, 5% nonfat dry milk, pH 7.5, and then incubated overnight with alkaline phosphatase-conjugated avidin
(BioRad, Richmond, CA). Bands were visualized using 5-bromo-4chloro-indoxyl phosphate and nitroblue tetrazolium (Sigma, St Louis,
MO) as substrate.40
Statistical analweJ. Each single clinical and phenotypic parameter has been matched with all of the others using the statistical technique of cross tabs. We used the Pearson x2 test (SPSS/PC) to define
the significance of each cross and analyzed the percentages in each
tab. When the probability value was less than .05, the hypothesis
that the two examined variables are independent was rejected.
The Student's t-test was used to evaluate statistical differences in
the mean lymphocyte counts among various phenotypic groups.
RESULTS
Seventy-four B-CLL patients were considered in this study.
In all of the samples, leukemic cells were enumerated via the
expression of HLA-DR and residual T cells were identified
through the expression of surface CD3 molecules.
In 52 of 74 patients, we determined percentages of HLADR' B-CLL cells greater than 70%. In the remaining patients,
these percentages ranged from 50% t o 70%. In the majority
of the patients (67 of 74), residual T cells were less than 20%.
Consequently, the threshold of positivity for each marker
was set at greater than 20%. In the 7 cases that had greater
than 20% T cells, the phenotypes of leukemic cells were assessed by two-color fluorescence analysis. No difference in
the fluorescence intensity was observed between freshly iso-
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
ADHESION MOLECULES ON B-CLL CELLS
268 1
lated and frozen cells, even for molecules, such as L-selectin,
that shed from the cell surface.''
Expression of CDS4 (ICAM-I) and CD58 (LFA-3) on BCLL cells. CD54 was detected on B-CLL cells from 13 of
74 patients; CD58 was present on cells from 6 of 13 CD54+
cases and was never found independently of CD54 expression.
In 2 of 3 patients whose cells expressed CD54, peripheral
blood and bone marrow samples were available simultaneously. In both cell samples, CD54 and CD58 were expressed
at similar proportions.
Expression of PI (CD29), p2 (CD18),and p3 (CD61) integrins on B-CLL cells. We analyzed the surface expression
of /3,, p2, and fi3 integrin chains on cells from 74 B-CLL
cases. In addition, we determined the expression of aL
(CDl la, LFA-I), aM(CD1 lb, Mac-I), and ax (CD1 IC,Leu
M5), the a chains known to associate with leukocyte (p2)
integrin~.~'.~'
Expression of PI, p2, and ,B3integrin chains on B-CLL cells
allowed us to distinguish three groups of B-CLL cases in which
cells did not express detectable surface PI, P2, and p3 chains;
or expressed PI and/or &, but not & integrin chains; or else
were positive for all three of the p chains (Table 2 and Fig
1).
Due to the sensitivity threshold of the indirect immunofluorescence technique routinely used to determine integrin
/3 chain expression and to assess whether a negative result
could indicate absence of a given p chain or its expression at
a very low density, we tested different detection systems. In
one approach, labeled antimouse Ig isotype (IgG2a and IgGI)
antibodies were used as secondary reagents to detect chain
expression. Fluorescence-activated cell sorter (FACS) analyses
did not show differences in fluorescence intensity when these
reagents were compared with secondary antitotal mouse Ig
antibodies (data not shown). In contrast, significant differences in /3 chain expression on B-CLL cells were shown when
results from indirect immunofluorescence assays were com-
pared with those obtained using biotinylated primary mouse
MoAb followed by fluorescein-labeled streptavidin (Fig 2).
While confirming previous results showing that, in most
B-CLL cases, PI and p2 integrin chains are expressed on the
surface of malignant B ~ e l l s , ~ ~we
~ ~observed
~ - ' " ' that, in 12 of
74 patients (Table 2), B-CLL cells also expressed surface p3
chains detectable by indirect immunofluorescence (Fig 1).
Lymphocytes from 4 patients, 3 positive for surface p3 chains
and 1 negative by FACS analysis, were labeled with biotin
and immunoprecipitated with anti-CD6 1 MoAb followed by
SDS-PAGE. A band of 90 Kd was present in the three cases
whose cells expressed p3 chains detectable by immunofluorescence. No band was detected in the p; case (Fig 3). An
unidentified band of approximately 68 Kd found in two cases
(lanes B and C in Fig 3 ) could be a proteolytic fragment of
the 90-Kd polypeptide. In all of the
cases we also found
coexpression of surface PI and p2 chains.
To rule out that detection of p3 chains could have been
due to platelets adherent nonspecifically to the malignant
cells, we stained B-CLL cells from three @: and three p; cases
with antibodies to the platelet-specific glycoprotein (GP) IIbIIIa (CD4la). Results were consistently negative. This finding
is also reinforced by the lack of staining of B-CLL cells from
all cases with MoAb to CD62 (P-selectin), a marker of activated platelets and endothelial cells.30 Finally, no adherent
platelets were detected in Giemsa-stained cytocentrifuge
preparations of B-CLL cells.
Analysis of PI and p2 chain expression allowed us to distinguish a group of patients whose cells did not express detectable PI and p2 chains. These cases (30 of 74) were also
negative for p3 chains. In 32 of 74 patients, PI (5 of 32), p 2
(9 of 32), or, more commonly, both chains (18 of 32) were
found on B-CLL cells at detectable amounts in the absence
of /33 chain expression (Table 2). In all of the cases, p2 chains
were coexpressed with aL chains to form the CD1 la/CD18
Table 2. lntegrin Chain Expression and Clinical Features a t the Time of Study
Patients
Sex (M/F)
Age (Median)
Rai stages 3-4 and Binet stages B-C
LDT <12 mo
Splenomegaly
Peripheral blood lymphocytes (x 1 03/pL)
Diffuse bone marrow pattern
Therapy requirement
12
30
18/12
62
9/30
6/26
11/30
32
17/15
61
10132
9/24
21/32
517
63.5
3/12
4/10
10112
28.9 f 23.0
44.8 k 44.0
30.9 f 26.4
3/22
612 1
511 1
10130
15/32
9/12
NS
NS
1 v 2 and 3 = .004
3 v 1 and 2 = ,042
2 v 1 and 3 = NS
1v2=NS
1 v3=NS
2v3=NS
1v2=NS
1 v 3 = ,044
2v3=NS
1 v 2 and 3 = NS
3 v 1 and 2 = .027
2 v 1 and 3 = NS
Expression of p integrin chains was determined by indirect immunofluorescence and flow cytometric analysis. Thus, negative means undetectable
using this technique. See text and Fig 2 for analysis and comment on these data.
Abbreviation: NS, nonsignificant.
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2682
DE ROSS1 ET AL
83.0
88.9
8.5
2.0
98 1
2.5
79.0
65
0.7
ik
i ',
.I
e-.
CD 61
(LFA-1) heterodimer; aM (CD1 lb, OKMI/MacI) and ax
(CD1 IC, LeuM5) were detected, respectively. on p i cells from
1 of 41 and 14 of 41 patients. Cells from the latter group of
patients were analyzed by two-color direct immunofluorescence to determine percentages of leukemic (HLA-DR') cells
also expressing LeuM5. Figure 4 shows three representative
cases in which HLA-DR+ cells also expressed LeuM5 at variable proportions. Notably, p3 was found in 6 of 14 cases in
which cells were LeuM5'. The difference in p3 expression on
cells from LeuM5' versus LeuM5- cases was highly significant (P= .0027).
In seven cases in which blood and bone marrow samples
were simultaneously available, we found that expression of
integrin @ chains was identical in cells from five patients. PI
chains were detectable on blood but not on bone marrow BCLL cells in one case. The opposite occurred in another case.
Expression ofL-selectin on B-CLL cells. L-selectin (Leu8)
was detectable on the surface of B-CLL cells in 2 1 of 74 cases.
Variable proportions of leukemic cells were positive for Lselectin in the various cases (Fig 5). We did not find significant
differences in the frequency of L-selectin-positive cases
among the patient groups expressing different sets of p chains.
In one of seven cases in which blood and marrow samples
80.0
331,
Fig 1 . Detection of B, ,&, and 8, integrin chains
on cells from three representative B-CLL cases.
Mouse MoAb anti+ chain was used, followed by
FITC-labeledgoat antimouse lg antibodies. Left row:
j3;, 0 2 , p; case. Middle row: a case in which most
cells express detectableB, and &, but not 8, chains.
Right row: a case positivefor all of the three chains
analyzed. HLA-DR and CD3 are used as markersfor
leukemic B cells and for residual T cells, respectively.
Percentagesof positivecells are reported at the top
right corner of each profile.
were tested simultaneously, L-selectin was detected on blood
but not on bone marrow cells. No difference was observed
in the remaining cases.
Expression qfCD44 on B-CLL cells. CD44 was detectable on B-CLL cells from all patients but one. The intensity
of CD44 expression on leukemic cells was compared with
that of residual T cells by two-color immunofluorescence.
As shown in Fig 6, CD44 was expressed on CD3' cells at
higher density than on HLA-DR+ (€3-CLL) cells. However,
in several cases, both cell types expressed CD44 at the same
density. This allowed us to distinguish three groups of patients
according to high (16 of 73), intermediate ( 1 8 of 73), or low
(39 of 73) CD44 expression. No difference was found when
CD44 expression was analyzed on blood and bone marrow
B-CLL cells.
Correlation between CD5 and adhesion molecule expression on B-CLL cells. The adhesion molecule profile was
evaluated in the 17 patients whose cells were CD5- versus
57 CD5' cases. We found an increased incidence (P= .049)
of CD5- cases in the LeuMS' group.
Correlation between adhesion molecule expression and
clinicalfeatures. Analysis of adhesion molecule expression
allowed for the distinction of subsets of patients whose cells
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
ADHESION MOLECULES ON B-CLL CELLS
B3 integrin chain expression on
the same B-CLL cell sample
using t w o different detection
systems. Top row: indirect immunofluorescence assay using
anti+'
chain mouse MoAb fol-
2683
11,,1
chain mouse MoAb followed by
FITC-labeled streptavidin. The
sensitivity of the avidin-biotin
assay is much higher than that
CD 18
cence test.
did or did not express the various types of adhesive receptors
at detectable levels. All of the phenotypic differences were
related to clinical parameters, namely, stage of disease according to Rai and Binet. LDT, peripheral blood lymphocytosis. splenomegaly. hepatomegaly, nodal involvement,
bone marrow infiltration pattern, and therapy requirement.
90-
A
B
C D
Fig 3. Immunoprecipitationof surface B3 chains from B-CLL cells
(see Materials and Methods for technical details). MoAbs to CD61
fail to precipitate B3 chains from cells of a 0; case, as determined
by immunofluorescence and flow cytometric analysis (lane A). A
90-Kd band is seen in lanes B, C, and D, which refer to 3 distinct
patients whose cells expressed B3 chains detectable by immunofluorescence. The band of 68 Kd seen in lanes B and C corresponds
presumably to a proteolytic fragment of the 90-Kd polypeptide.
Patients whose cells expressed CD58 were more frequently
assigned to the Rai 3 and 4 and Binet B and C stages (P=
.038) and exhibited a higher incidence of splenomegaly (P=
. O B ) . As for CD54. we did not find significant correlations
with the evaluated clinical parameters.
Several correlations between B-CLL cell phenotype and
clinical features were found when integrin p chain expression
was analyzed (Table 2). Cases lacking P I , p2, and & chain
expression had a significantly lower incidence ofsplenomegaly
(P = .004). When cases lacking all of the analyzed p chains
were compared with those expressing all three p chains, it
was found that the former had a significantly lower incidence
of diffuse bone marrow infiltration (P = .044) and oftherapy
requirement (P = .027) (Table 2). Among the three groups.
p chain-negative cases had the lowest peripheral blood lymphocytosis. but these data were not statistically significant (P
= .08).
In view of the possibility that (Y& (LeuM5) expression
distinguishes a subset of patients with a more aggressive disease ~ o u r s e . o~r ~variants
. ~ ~ of chronic lymphoproliferative
disorders?5 we examined the clinical features of 14 LeuM5'
versus 60 LeuM5- cases (Table 3). Patients in the former
group presented with a significantly higher lymphocytosis (P
= .021) and need for therapy (P = .006).
L-selectin was detected on cells from 21 of 74 cases. No
significant correlation has been observed between clinical parameters and Leu8 expression.
As for the variable intensity ofCD44 expression on B-CLL
cells. the only clinical difference among the three groups was
that between CD44'"w and CD44highcases. A significantly increased incidence of diffuse bone marrow infiltration was
found in the latter group (P = .O 12).
DISCUSSION
B-CLL is a disorder with a high degree of biologic and
clinical heterogeneity.'9.2n.2xSurface molecules mediating cell-
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2684
'r
C I -
DE ROSS1 ET AL
17
-5!
l-
51
81
'
2
DH
Fig 4. Two-color immunofluorescence analysis of HLA-DR and ax& chain (LeuM5) coexpression on B-CLL cells from three distinct
patients. Variable percentages of double-positive cells are determined in each case (percentage values are reported in square 2). Doublenegative cells in square 3 correspond to residual T cells.
to-cell and cell-to-ECM adhesion could be good candidates
for phenotypic to clinical correlations in view of their role
for tumor cell spreading and m e t a ~ t a s i s . ~ ~
Expression of adhesion molecules may also explain certain
functional characteristics of B-CLL cells, such as their ability
to adhere to substrates. A previous study has shown that BCLL cells express PI and pz integrin chains and that these
molecules codistribute with cytoskeletal proteins at specific
adhesion sites called p o d o ~ o m e s In
. ~ the
~ same study, B3 integrin chains were undetectable and, quite intriguingly, no
a chains could be shown in association with PI and Pz chains.
This raises the important issue of the detection of surface
adhesion molecules. Experiments shown in Fig 2 clearly
demonstrate that assessment of surface integrin expression
(as well as that of other adhesion molecules) on B-CLL cells
largely rests upon the detection system that is used. Therefore,
negativity or positivity for surface adhesion molecules is of
relative value, and is probably attributable to the surface
density of any given molecule rather than to its presence or
absence. This may hamper a possible comparison of phenotypic analyses from different laboratories and may explain
some conflicting data reported in the literature. However, we
believe that our classification of B-CLL cases into phenotypic
subsets (as shown, for example, in Table 2) reflects true phenotypic differences and that it lends itself to the possibility
of clinical correlations provided that cells from all cases are
analyzed using the same reagents and the same detection
system.
Adhesion molecules of the Ig superfamily (CD54 and
CD58) were found in a minority of B-CLL cases. As for other
cell types (eg, endothelial cells and keratinocytes), this may
reflect a state of cell activation induced by cytokines produced
by the malignant B cells or by other cell types.47
While confirming previous studies describing the expression of PI and p2 integrin chains on B-CLL ce11s,23.42-44
we
found that cells from 12 of 74 B-CLL patients also expressed
p3 chains, always in association with PI and p2 chains. This
novel finding was confirmed by both flow cytometric and
Fig 5. Two-color immunofluorescence analysis of HLA-DR and L-selectin(Leu8) coexpressionon B-CLL cells from three distinct patients.
Variable percentages of double-positivecells are shown in the three cases (percentage values are reported in square 2). Double-negative
cells in square 3 correspond to residual T cells.
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
ADHESION MOLECULES ON B-CLL CELLS
2685
biochemical analyses and we could rule out that it was due
to adherent platelets (see Results). Therefore, we could define
distinct groups of patients according to the expression of (3,,
Pz, and & integrin chains.
No major differences were noted when expression of the
above described adhesion molecules was analyzed on B-CLL
cells simultaneously obtained from blood and bone marrow
or when CD5' clones were compared with CD5- clones, with
the exception of a significantly higher incidence of LeuMS
clones in the latter group.
The high degree of biologic and clinical heterogeneity of
B-CLL has prompted numerous studies aimed at the detection of risk factors. These studies have been addressed both
at the definition of phenotypic heterogeneity of B-CLL cells
and of several clinical parameters. We selected groups of patients with distinctive adhesion molecule phenotypes. These
groups were evaluated in relation to a variety of clinical features. Staging according to Rai and to Binet, LTD, and the
bone marrow histologic pattern have been considered important prognostic parameters in survival analyses of BCLL.35-38
Furthermore, therapy requirement is related to the
appearance of negative risk factors and/or to disease progression. Therefore, we considered the need for therapy as a
relevant prognostic variable. Evaluation of survival time could
not be taken into account because at least one-half of the
patients had a follow-up period shorter than 3 years.
The main conclusionsthat can be drawn from a correlation
between adhesion molecule phenotype and clinical features
are the following. Patients whose cells do not express integrin
p chains (PI, &, and b3) show the most favorable clinical
Table 3. CD11 c/CD18 (LeuM5) Expression and Clinical Features
at the Time of Study
Patients
Rai stages 3-4 and
Binet stages B-C
LTD 1 1 2 mo
Splenomegaly
Peripheral blood lymphocytes
(X 103IrL)
Diffuse bone marrow pattern
Therapy requirement
LeuM5-
LeuM5+
60
14
15/60
15/49
31/60
6/14
4/13
11/14
NS
NS
NS
31.6 f 28.8
9/44
23/60
55.3 i 49.6
5/10
11/14
,021
NS
P
,006
Abbreviation: NS, nonsignificant.
features as compared with the other groups (Table 2).
Expression of ax& integrin (LeuMS) has allowed for the detection of a B-CLL subset with clinical features similar to
those of HCL and of other types of B-CLL lymphoproliferative disorder^.^^,^^,^^ In the 14 LeuMS+ cases of our series,
we did not find most of the features described in the above
studies. However, patients whose malignant cells expressed
LeuM5 differed from LeuMS- cases for a significant incidence
of higher lymphocytosis and of therapy requirement (Table
3).
Furthermore, expression of CD58 was detected in cases
with higher incidence of splenomegaly and advanced Rail
Binet stages. Diffuse bone marrow infiltration was significantly more frequent in patients whose cells expressed
CD44high.This is consistent with the finding that CD44 me-
&
n
CD 44
.
1
2 .
77
" ?
Fig 6. Expression of CD44 on B-CLL cells and
on residual T cells from two distinct patients. B-CLL
cells were identified as HLA-DR+, whereas T cells
were labeled with MoAb to CD3. Top row: B-CLL
cells (left] and T cells (right) express CD44 at the
same intensity. Bottom row: B-CLL cells express
CD44 at lower density in comparison to that determined on T cells. Percentages of double-positive
cells are reported at the top right corner.
.
:@
:
m -
"Id'
IYI
in'
CD 44
10'1
194
101
CD 44
IO*
I.'
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
2686
DE ROSS1 ET AL
diates adhesion of maturing B cells to bone marrow stromal
cells.48
Analyses of the relationships of clinical features with cell
phenotypes have their major constraint in the long median
survival of B-CLL patients. However, LDT and therapy requirement are evidently dynamic variables and will therefore
be evaluated in the long run.
ACKNOWLEDGMENT
We are most grateful to Dr Piero Mancini (Dept. of Statistics,
University La Sapienza, Rome, Italy) for statistical analyses. We thank
Giuliana Buscaglia and Maura Seven for their secretarial assistance.
We also thank Dr K. R. Rai for critical reading of the manuscript
and for helpful suggestions.
REFERENCES
1. Greaves MF: Differentiation-linked leukemogenesis in lymphocytes. Science 234:697, 1986
2. Rusciano D, Burger MM: Why do cancer cells metastasize into
particular organs? Bioessays 14: 185, I992
3. Albelda SM, Buck CA: Integrins and other cell adhesion molecules. FASEB J 4:2868, 1990
4. Moller G (ed): Adhesion molecules. Immunol Rev 1145, 1990
5. Shimizu Y, Shaw S: Lymphocyte interactions with extracellular
matrix. FASEB J 5:2292, 1991
6. Maio M, Pinto A, Carbone A, Zagonel V, Gloghini A, Marotta
G, Cirillo D, Colombatti A, Ferrara F, Del Vecchio L, Ferrone S:
Differential expression of CD54/intercellular adhesion molecule I in
myeloid leukemias and in lymphoproliferative disorders. Blood 76:
783, 1990
7. Stander R, Grail R, Schultz TF, Thaler J, Gattringer C, Radaskiewicz T, Dierech MP, Huber H: Expression of leucocyte-function
associated antigen- 1 and 7F7 antigen, an adhesion molecule related
to intercellular adhesion molecule- I (ICAM-I) in non-Hodgkin lymphomas and leukemias: Possible influence on growth pattern and
leukaemic behaviour. Clin Exp lmmunol 77:234, 1989
8. Medeiros LJ, Weiss LM, Picker W, Clayberger C, Hornung SJ,
Krensky AM, Warnke RA: Expression of LFA-I in non-Hodgkin
lymphoma. Cancer 63:255, 1989
9. Horst E, Radaskiewicz T, Hoofman-den Otter A, Pieters R,
Van Dongen JJM, MelJerCJLM, Pals ST: Expression of the leucocyte
integrin LFA-I (CDI Ia/CD18) and its ligand ICAM-1 (CD54) in
lymphoid malignancies is related to lineage derivation and stage of
differentiation but not to tumor grade. Leukemia 55348, 199 I
IO. Horst E, Meijer CJLM, Radaskiewicz T, OssekoppeleGJ, van
Krieken JHJM, Pals ST: Adhesion molecules in the prognosis of
diffuse large-cell lymphoma: Expression of a lymphocyte homing
receptor (CD44), LFA-I (CDI la/18), and ICAM-1 (CD54). Leukemia
8595, 1990
I 1. Spertini 0, Freedman AS, Belvin MP, Penta AC, Griffin JD,
Tedder TF: Regulation of leukocyte adhesion molecule-1 (TQ 1, Leu8)
expression and shedding by normal and malignant cells. Leukemia
5:300, 1991
12. Michie SA, Garcia CF, Strickler JG, Dailey M, Rouse RV,
Warnke RA: Expression of the Leu8 antigen by B-cell lymphomas.
Am J Pathol 88:486, 1987
13. Carbone A, Manconi R, Poletti A, Gloghini A, De Paoli P,
Volpe R Expression of Leu8 surface antigen in €3-cell lymphomas.
Correlation with other B-cell markers. J Pathol 154:133, 1988
14. Pals ST, Horst E, Ossekoppele GJ, Figdor CG, Scheper RJ,
Meijer CJLM: Expression of lymphocyte homing receptor as a
mechanism of dissemination in non-Hodgkin lymphoma. Blood 73:
885, 1989
15. Horst E, Meijer CJLM, Radaskiewicz T, van Dongen JJM,
Pieters R, Figdor CG, Hooftman A, Pals ST: Expression of a human
homing receptor (CD44) in lymphoid malignancies and related stages
of lymphoid development. Leukemia 4:383, I990
16. Picker LJ, Medeiros U,Weiss LM, Warnke RA, Butcher EC:
Expression of lymphocyte homing receptor antigen in non-Hod&n’s
lymphoma. Am J Pathol 130:496, 1988
17. Van Riet I, De Waele M, Remels L, Lakor P, Schots R, Van
Camp B: Expression of cytoadhesion molecules (CD56, CD54, CDI 8
and CD29) by myeloma plasma cells. Br J Haematol 79:42 I , 1991
18. Ashmann EJM, Lokhorst HM, Dekker AW, Bloem AC:
Lymphocyte function-associated antigen-I expression on plasma cells
correlates with tumor growth in multiple myeloma. Blood 79:2068,
I992
19. Dighiero G, Travade P, Chevret S, Ferraux P, Chastang C,
Binet JL, and the French Cooperative Group: B-cell chronic lymphocytic leukemia: Present status and future directions. Blood 78:
1901, 1991
20. Geisler CH, Larsen JK, Hansen NE, Hansen MM, Christensen
BE, Lund B, Nielsen H, Plesner T, Thorling K, Andersen E, Andersen
PK: Prognostic importance of flow cytometric immunophenotyping
of 540 consecutive patients with B-cell chronic lymphocytic leukemia.
Blood 78:1795, 1991
21. Kipps TJ: The CD5 B cell. Adv Immunol 47:117, 1989
22. Caligaris-CappioF, Gobbi M, Boffil M, Janossy G: Infrequent
normal B lymphocytes express features of B-chronic lymphocytic
leukemia. J Exp Med 155:623, 1982
23. Morabito F, Prasthofer EF, Dunlap NE, Grossi CE, Tilden
AB: Expression of myelomonocytic antigens on chronic lymphocytic
leukemia B cells correlates with their ability to produce interleukin1. Blood 7 0 1750, 1987
24. De La Hera A, Alvares-Mon M, Sanchez-Madrid F, Martinez
AC, Durantez A: Co-expression of Mac- 1 and pl50,95 on CD5+ B
cells. Structural and functional characterization in a human chronic
lymphocytic leukemia. Eur J Immunol 18:1 13I , 1988
25. Hanson CA, Gribbin TE, Schnitzer B, Schlegelmilch JA,
Mitchell BS, Stoolman LM: CDI IC (LeuMS)expression characterizes
a B-cell chronic lymphoproliferative disorder with features of both
chronic lymphocytic leukemia and hairy cell leukemia. Blood 76:
2360, 1990
26. Wormsley SB. Baird SM, Gadol H, Rai KR, Sobol RE: Characteristics of CDI Ic+CD5+ chronic B-cell leukemias and the identification of novel peripheral blood B-cell subsets with chronic lymphoid leukemia immunophenotype. Blood 76: 123, I990
27. Grossi CE, Zicca A, Leprini A, Cadoni A, Pistoia V, Ferrarini
M: Acid hydrolases as markers of maturation in B-cell chronic lymphocytic leukemia. Blood 60:220, I982
28. Cheson BD, Bennet JM, Rai KR, Grever MR, Kay NE, Schiffer
CA, Oken MM, Keating MJ, Boldt DH, Kempin ST, Foon KA:
Guidelines for clinical protocols for chronic lymphocytic leukemia.
Recommendations of the National Cancer Institute-sponsored
Working Group. Am J Hematol 25:152, 1988
29. Springer TA, Dustin ML, Kishimoto TK, Marlin SD: The
lymphocyte function-associated LFA- I , CD2, and LFA-3 molecules:
Cell adhesion receptors of the immune system. Ann Rev Immunol
5:223, 1987
30. Springer TA: Adhesion receptors of the immune system. Nature 346:425, 1990
3 1. Ruoslahti E: Integrins. J Clin Invest 87: 1, 199I
32. Butcher EC: Leukocyte-endothelial cell recognition: Three (or
more) steps to specificity and diversity. Cell 67:1033, 1991
33. Varki A: Selectins and other mammalian sialic acid-binding
lectins. Curr Opin Cell Biol 4:257, 1992
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
ADHESION MOLECULES ON B-CLL CELLS
34. Hynes RO: Integrins: Versatility, modulation and signaling in
cell adhesion. Cell 69: 1 I , 1992
35. Rai KR, Sawitsky A, Cronkite EP, Chanana AD, Levy RN,
Pasternak BS: Clinical stagingof chronic lymphocyticleukemia. Blood
46:219, 1975
36. Binet JC, Catovsky D, Chandra M, Montserrat E, Rai KR,
Sawitsky A (Writing Committee of the International Workshop on
CLL): Chronic lymphocytic leukemia: Proposal for a revised staging
system. Br J Haematol 48:365, 1981
37. Montserrat E, Sanchez-Bison0 J, Vinolas N, Rozman C:
Lymphocyte doubling time in chronic lymphocytic leukemia:
Analysis and its prognostic significance. Br J Haematol 62567,
1986
38. Rozman C, Montserrat E, Rodriguez-Fernandez JM, Ayats
R, Vallespi T, Parody R, Rios A, Prado D, Morey M, Gomis F,
Alcala A, Gutierrez M, Maldonado J, Gonzalez C, Giralt M, Hernandez-Nieto L, Cabrera A, Fernandez-Ranada JM: Bone marrow
histologic pattern. The best single prognostic parameter in chronic
lymphocytic leukemia: A multivariate survival analysis of 329 cases.
Blood 64:642, 1984
39. Von Boxberg Y, Wutz R, Schwarz U: Use ofthe biotin-avidin
system for labelling, isolation and characterization of neural cell surface proteins. Eur J Biochem 190:175, 1990
40. Blake ML, Johnson KH, Russel-Jones GJ, Gotschlich EC: A
rapid and sensitive method for detection of alkaline phosphatase
conjugated anti-antibody on Western blots. Anal Biochem 136:175,
1984
2687
41. Hemler ME: VLA proteins in the integnn family: Structures,
functions, and their role on leukocytes. Ann Rev lmmunol 8:365,
I990
42. Marchisio PC, Bergui L, Corbascio GC, Cremona 0, DUrso
N, Schena M, Tesio L, Caligaris-Cappio F Vinculin, talin, and integrins are localized at specific adhesion sites of malignant B lymphocytes. Blood 722330, 1988
43. Caligaris-Cappio F, Bergui I, Schena M, Chilosi M, Marchisio
PC: Cell-cell and cell-matrix adhesion structures may influence the
growth pattern of chronic lymphoid malignancies. Leukemia 3: 167,
1989
44. Baldini L, Cro L, Calori R, Nobili L, Silvestri I, Maiolo A T
Differential expression of very late activation antigen-3 (VLA-3)/VLA4 in B-cell non-Hodgkin lymphoma and B-cell chronic lymphocytic
leukemia. Blood 79:2688, 1992
45. Bassan R, Neonato MG, Abbate M, Motta T, Barbui T, Rambaldi A: Monoclonal lymphocytosis with villous lymphocytes: A
chronic lymphoproliferative disease of CD 1 IC' B-cells. Leukemia 5 :
799, I99 I
46. Nicolson GL: Organ specificity of tumor metastasis: Role of
preferential adhesion, invasion and growth of malignant cells at specific secondary sites. Cancer Metastasis Rev 7:143, 1988
47. Dejana E, Bertocchi F, Bortolami MC, Regonesi A, Tonta A,
Breviario F, Giovazzi R: Interleukin 1 promotes tumor cell adhesion
to cultured human endothelial cells. J Clin Invest 82:1466, 1988
48. Miyake K, Underhill CB, Lesley J, Kincade PW: Hyaluronate
can function as a cell adhesion molecule and CD44 participates in
hyaluronate recognition. J Exp Med 172:69, 1990
From www.bloodjournal.org by guest on June 18, 2017. For personal use only.
1993 81: 2679-2687
Adhesion molecule expression on B-cell chronic lymphocytic
leukemia cells: malignant cell phenotypes define distinct disease
subsets
G De Rossi, D Zarcone, F Mauro, G Cerruti, C Tenca, A Puccetti, F Mandelli and CE Grossi
Updated information and services can be found at:
http://www.bloodjournal.org/content/81/10/2679.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.