[CANCER RESEARCH 41, 4050-4056. October 1981)
0008-5472/81/0041-0000802.00
Comparison of Lipid Composition and 1,6-Diphenyl-1,3,5-hexatriene
Fluorescence Polarization Measurements of Hairy Cells with
Monocytes and Lymphocytes from Normal Subjects and Patients
with Chronic Lymphocytic Leukemia1
Leonard F. Liebes,2 Edward Pelle, Dorothea Zucker-Franklin,
and Robert Silber3
Department of Medicine. New York University Medical Center, New York. New York 10016
ABSTRACT
In this report, we compare the lipid composition and fluores
cence polarization properties of hairy cells with those of monocytes and lymphocytes from normal subjects and of lympho
cytes from patients with chronic lymphocytic leukemia. For
hairy cells, the cholesterol content was 4.66 ± 1.49 (S.D.)
jumol/109 cells, and the cholesterol/phospholipid
ratio was
0.60 ±0.09. These were significantly higher than the values
of normal lymphocytes, (cholesterol content, 2.75 ± 0.65
jumol; cholesterol/phospholipid
ratio, 0.50 ± 0.07) or of
chronic lymphocytic leukemia lymphocytes (cholesterol con
tent, 1.76 ±0.43 jumol; cholesterol/phospholipid
ratio, 0.44
±0.07). Normal monocyte values (cholesterol content, 5.81
± 2.08 /imol; cholesterol/phospholipid
ratio, 0.59 ± 0.06)
were similar to those of hairy cells. Using the probe 1,6diphenyl-1,3,5-hexatriene,
the fluorescence polarization value
at 25° for hairy cells was 0.302, compared to the value of
0.259 obtained with chronic lymphocytic leukemia lympho
cytes. Intermediate values (0.294) were obtained with normal
lymphocytes and monocytes. Fluorescence polarization values
were higher in hairy cell membranes than in chronic lympho
cytic leukemia lymphocyte membranes, indicating a low fluidity
in the former cell, compatible with their higher cholesterol
content and cholesterol/phospholipid
ratio. These results show
that two neoplastic cells, hairy cells and chronic lymphocytic
leukemia lymphocytes, differ markedly in membrane fluidity
and that a high membrane fluidity does not necessarily occur
in neoplasia.
INTRODUCTION
HCL4 is characterized
by the presence of circulating
mono-
nuclear cells of unique morphology (41). As indicated by their
name, HC have long processes that cover their surfaces (Fig.
1). The cell has mainly been considered to be a B-lymphocyte
(7, 13, 15, 23, 46) but also a monocyte or hybrid cell (22, 27,
33, 40). Large numbers of HC are found in the spleen of
patients with HCL (4). While the precise characteristics respon
sible for the retention of a cell by the splenic sinusoids remain
undefined, antibody coating, cell volume, and cell membrane
fluidity (9) have been considered as factors influencing the
ability of a cell to traverse this organ.
Since the peculiar configuration of the HC surface suggested
the possibility of either excessive or an anomalous plasma
membrane, a study was undertaken to compare the lipid com
position and the membrane fluidity of HC with those of mono
cytes and lymphocytes from normal subjects and with those of
lymphocytes from patients with CLL. Unlike the CLL lympho
cytes, the HC was found to be a lipid-rich neoplastic cell with
low membrane fluidity.
MATERIALS
AND METHODS
Clinical Data, Cell Isolation, and Membrane Preparation. Mononuclear cells were purified from 10 to 500 ml of blood obtained from
normal volunteers and from untreated patients with HCL and CLL. For
most experiments, heparin was used as an anticoagulant. As indicated
below, in our hands, identical results were obtained with heparin as in
the absence of anticoagulants, i.e., when defibrinated blood was used
as the source of mononuclear cells. Normal lymphocytes or monocytes
were obtained as the by-product of plateletpheresis on a cell separator
(42) using acid/citrate/dextrose
Formula B. All donors gave informed
consent according to the provisions of the Helsinki conference. The
clinical diagnosis of HCL was substantiated by morphological appear
ance and a positive tartrate-resistant acid phosphatase reaction of the
characteristic cells. The percentage of T- and B-cells determined by
sheep erythrocyte-forming
and erythrocyte-antibody-complementforming rosettes ranged from 3.5 to 10% and from 17 to 23% of the
mononuclear cells, respectively, in the patients with HCL. All the
patients studied had B-cell CLL with 48 to 98% of lymphocytes positive
for erythrocyte-antibody-complement
receptors (43). Mononuclear
cells were isolated by centrifugation on Ficoll-Hypaque gradients (29).
Normal monocytes were removed by adherence to plastic as described
previously (36, 38). The purity of cellular preparations which exceeded
90% for all experiments was ascertained by cell-sizing analyses on a
Model B Coulter Counter equipped with a 1OO-jim aperture and cali
brated with 10-ftm latex particles. Mononuclear cells from patients with
HCL consisted of 2 populations, a major one which was comparable to
monocytes in volume and a minor one slightly larger than normal and
CLL lymphocytes (36). All studies with HCL were carried out in prep
arations that contained from 65 to 95% HC. The T-cells were under
10% of the total mononuclear population in all cases, compatible with
the diagnosis of "B-cell type HCL." The ages of the HC patients ranged
' This study was supported by NIH Grants CA11655 and AM 12274.
2 Scholar of the Leukemia Society of America, Inc.
3 To whom requests for reprints should be addressed, at Department
of
Medicine. New York University Medical Center, 560 First Avenue, New York,
N. Y. 10016.
4 The abbreviations used are: HCL, hairy cell leukemia; HC, hairy cell(s); CLL,
chronic lymphocytic leukemia; C/PL ratio, cholesterol/phospholipid
1.6-diphenyl-1,3,5-hexatriene.
Received March 2. 1981 ; accepted July 7, 1981.
4050
ratio; DPH,
from 51 to 66 years; these overlap the CLL population ages of 45 to
70 years. These patients were not sex matched since all the patients
with HCL were males. The presence or absence of the spleen had no
effect on the HC cholesterol-phospholipid
content. While 6 of the 8
HCL patients had undergone splenectomy, no significant differences
were found in the cell lipid composition between the 2 groups [presplenectomy C/PL,
0.58 ±0.04 (S.D.); postsplenectomy
CANCER
C/PL,
RESEARCH
0.61 ±
VOL. 41
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
Lipids and Fluidity of HC, Monocytes,
and Lymphocytes
to dryness under N? and immediately subjected to a 1-hr acid hydrolysis
at 180°with the addition of 25 p\ 36 N H2SO„
and 100 /il H2O. With the
final volume brought up to 400 ¡i\with H?O, the phosphorus determi
nation was carried out according to the method of Chen ef al. (6).
Comparison of Lipid Assay Results in Lymphocytes Prepared
from Defibrinated or Heparinized Blood. The Ficoll-Hypaque mononuclear layer from the blood of normal subjects has a higher platelet/
lymphocyte ratio than did that from patients with CLL or HCL. Although
platelets removed from the standard heparinized blood preparation by
6 low-speed centrifugations of the mononuclear cells yielded normal
lymphocyte preparations containing fewer than one platelet per 5
lymphocytes, the effect of even more rigorous platelet depletion was
evaluated as follows. In 9 experiments, blood from normal subjects was
defibrinated rather than heparinized, yielding platelet-free preparations
Fig. 1. Electron micrograph
processes, x 10,000.
of a HC illustrates
the abundance
of cellular
0.10]. One HCL patient whose cellular lipid composition had been
studied 8 times prior to and twice postsplenectomy showed no change.
Lymphocytes from 2 normal donors who had been splenectomized had
a lipid composition comparable to that of lymphocytes from nonsplenectomized donors. Cells obtained from 3 patients with CLL who
manifested splenomegaly comparable to that found in HCL were no
different in lipid composition than were lymphocytes from the other
CLL patients. No subfractionation
of normal lymphocytes into T-enriched and B-enriched populations was done for 2 reasons: (a) studies
akin to the CLL specimens.
The molar C/PL ratio of the lymphocytes from these 9 subjects
(0.50) was identical to that obtained with lymphocytes from 19 normal
subjects (0.50) where heparin had been used as an anticoagulant. For
this reason, data on heparinized and defibrinated preparations were
pooled in subsequent experiments.
Fluorescence Polarization of Cells and Plasma Membranes. The
fluorescent probe DPH (Aldrich Chemical Co.) was used to monitor the
degree of fluorescence anisotropy (polarization) in intact cells and
isolated plasma membrane preparations. The fluorofluor was recrystallized in chloroform and stored at -10° in the dark. It was found to
be chromatographically
liquid chromatography.
homogeneous by reverse-phase high-pressure
For labeling of cells or membranes, 2 x 10~3
M DPH in tetrahydrofuran
pH 7.5, to a concentration
was diluted with phosphate-buffered
saline,
of 2 x 10~6 M with vigorous stirring at room
temperature. Whole cells or membranes were incubated at a concen
tration of 10~6 M probe per 100 ¡igprotein for 30 min at 37°with gentle
shaking. Measurements were obtained over a concentration range of
15 to 40 fig/ml protein to correct for light scattering. Steady-state
polarization values were obtained on either an Elscint MV-1 microviscometer (Elscint Corp., Hackensack, N. J.) or a Hitachi Perkin-Elmer
by others have shown that these cells have similar fluorescence polar
ization (P) properties (25, 50; and to) the methods for separating these
cells by either rosette formation or immunoabsorbance may well affect
membrane properties since they involve ligand binding. For lipid and
fluorescence
polarization determinations,
plasma membranes were
prepared from 2 to 10 x 108 HC, monocytes, normal lymphocytes,
MPF3 spectrofluorometer
equipped with a Wood modification allowing
automatic positioning of the polarization filters (C. M. Wood Manufac
turing Co., New Town, Pa.). Temperatures on the Perkin-Elmer spec
trofluorometer were maintained with a Perkin-Elmer digital temperature
and CLL lymphocytes by a technique described previously (28). The
purification was monitored with 5'-nucleotidase
assays and electron
degree of steady-state
microscopy.
Whole-cell
and plasma membrane protein was measured
by the procedure of Lowry ef al. (30) or the fluorescamine method (2)
using bovine serum albumin as a standard.
Extraction of Lipids. An aqueous Folch extraction (14) was carried
out on purified whole cells and membranes as follows. For extraction
of whole cells, fresh pellets containing 108 cells were resuspended in
0.5 ml methanol. After the addition of 1 ml chloroform, the mixture was
agitated for 2 min. A minimum time of 1 hr at room temperature was
allowed prior to the addition of an equal volume of 0.06 N HCI. For
membrane lipid determinations,
aliquots of membrane preparations
containing 100 to 200 ^g protein were suspended at a protein concen
tration of 0.3 mg/ml in 0.010 M Tris buffer, pH 7.0, and mixed with 0.5
ml methanol followed by 1 ml chloroform. Further processing was as
described for whole cells. The chloroform layer was removed for assay
of the cholesterol and phospholipid.
Assay for Total Cholesterol. The analysis for whole-cell and mem
brane cholesterol routinely used the colorimetrie determination of Rudel
and Morris (39). Because this may give falsely high estimates of
cholesterol (24), membrane and whole-cell cholesterol determinations
were also performed using the cholesterol oxidase enzyme assay (1,
19). Cholesterol 99 + purity, o-phthalaldehyde,
cholesterol oxidase
(20.0 units/ml), horseradish peroxidase (275 units/ml), and p-hydroxyphenylacetic acid were obtained from Sigma.
Phospholipid Assay. Aliquots of the chloroform extract were taken
OCTOBER 1981
controller, coupled to a thermoelectric
cell holder. Further details
concerning these instruments are provided in the "Appendix."
The
fluorescence
anisotropy
expressed
as polari
zation (P value) is used in this study as an expression of the degree of
freedom of the DPH molecules to rotate in the lipid environment in
which they are dissolved. For the purposes of this study, the range of
P values is used as a scale of relative fluidity. It was felt that conversion
to a measurement of absolute microviscosity utilizing the Perrin equa
tion was not appropriate.
Ultrastructure Studies. Electron microscopic analyses of HC were
performed by previously described methods (51 ). The membrane frac
tions isolated from monocytes and HC were prepared essentially as
described elsewhere (28). They were centrifuged into a pellet at
100,000 x g for 1 hr, fixed in 3% phosphate-buffered
glutaraldehyde
for 1 hr or overnight, and postfixed with 2% OsCX for 2 hr. Dehydration
was accomplished with increasing concentrations of alcohol and propylene oxide. The embedding medium was Epon 812. Thin sections
were contrasted with uranyl acetate and lead citrate. The preparations
were surveyed for purity at original magnifications of x 15,000 and
x30,000 with a Siemens Elmiskop I electron microscope at an accel
erating voltage of 60 kV.
RESULTS
Cholesterol-Phospholipid Composition of Whole Cells and
Membranes. The lipid composition of intact HC differed from
that of normal and CLL lymphocytes in several ways (Table 1).
The major finding was a significantly higher cholesterol content
4051
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
L. F. Liebes et al.
Table 1
Whole-cell cholesterol and phospholipid
values
Test of significance (Student's ( test for comparison of population means of all data expressed in terms of ftmol/109
cells): HC versus ML," CLL (C, PL), p £0.003; CLL versus NL, MONO (C, PL), p < 0.001 ; NL versus MONO (C, PL),
p s 0.003; HC versus MONO (C, PL), p > 0.17. Mann-Whitney U test using critical values for a one-tailed test: HC
versus NL, CLL (C/PL), p < 0.01 ; CLL versus NL, MONO (C/PL), p < 0.01 ; NL versus MONO (C/PL), p < 0.01 ; HC
versus MONO (C/PL), not significant.
Cholesterolfimol/109
pro
cellsHCMonocyte
tein82.1
1.496
5.812.75
2.08±
±
± 8.9
18.861.4
87.7 ±
cells2.87tein151
(8)c
±144
10.05.56
3.931.20±121.4 26.522.520.0C/PL0.60
0.590.50
±±
±±
(9)0.07
0.06
Lymphocytes
Normal
±103.7
(23)
0.65
±13.1
1.76±
±0.43nmol/mg
48.3 ±11.5/imol/10s7.97
3.95±
0.81nmol/mg ±pro-19.1 0.44±
0.07 (24)
CLL4.66
±0.09
±Phospholipid'
a NL. normal lymphocytes; C, cholesterol; PL, phospholipid; MONO, monocytes.
6 Mean ±S.D.
c Numbers in parentheses, number of determinations.
and C/PL ratio observed in the HC when compared to normal
or CLL lymphocytes, indicating the selective enrichment of HC
with cholesterol. Consistent with their greater volume (3), HC
had higher contents of cholesterol and phospholipid per cell.
The concentration of these lipids in HC, however, was also
higher when expressed per mg of cell protein. The results in
Table 1 also show that the cholesterol and phospholipid content
of HC was similar to that of normal monocytes. The lowest
cholesterol content when expressed on a per cell basis and C/
PL ratio was found in CLL, with intermediate values in normal
lymphocytes. When expressed per mg of cell protein, the
difference between CLL and normal lymphocytes was de
creased since the mean protein content of CLL lymphocytes
was 41.5 mg/109 cells compared to 44.8 mg/109 cells for
normal lymphocytes. The results reported above were found
with either the enzymatic or the o-phthalaldehyde assay for the
determination of cholesterol. The cholesterol oxidase assay
yielded levels which were about 10% lower than those of the
other assay, regardless of the cell type in which cholesterol
was being measured (Table 2).
The lipid composition of plasma membranes prepared from
HC, lymphocytes and monocytes was also investigated. The
method used for preparing membranes has been described
previously for normal and CLL lymphocytes (28). Its applica
bility to normal monocytes and HC was validated by the follow
ing: (a) these plasma membrane preparations were enriched
10- to 20-fold in 5'-nucleotidase activity; (b) the preparations
were devoid of nuclei and mitochondria as judged by electron
microscopy. Although some nonmembrane material was pres
ent in all specimens, the amount of these contaminants was
similar in membrane fractions prepared from normal monocytes
and HC (compare Figs. 2/4 and 2B). (c) The cholesterol and
phospholipid content of plasma membrane preparations ob
tained by this method show the usual enrichment in lipid/
protein ratio over the cells from which they originated (Tables
1 and 3). The data presented in Table 3 and Chart 1/4 also
show that HC membranes had a slightly higher cholesterol
content expressed per mg protein than did the membranes
prepared from normal or CLL lymphocytes. In addition, Chart
1,4 also shows that membranes of normal lymphocytes, mon
ocytes, and CLL lymphocytes had similar C/PL composition.
Fluorescence Polarization Values of Cells and Plasma
Membranes. Polarization measurements using DPH as a probe
were obtained on whole-cell preparations at 25°and 37°. HC
4052
Table 2
Comparison of 2 whole-cell cholesterol determinations
oxide
(nmol/
cells)4.01
10*
(/imol/109
cells)4.47
1.19°2.48
±
±1.112.27
Lymphocytes
±0.41
±0.49
Normal (6)
1.79 ±0.24Cholesterol1.60 ±0.25%
CLL (9)o-Phthalaldehyde
' (Cholesterol oxidase/o-phthalaldehyde) x 100%.
' Numbers in parentheses, number of samples.
: Mean ±S.D.
of differ
ence889.791.5
HC (5)"
89.0
Table 3
Lipid composition
of mononuclear cell membranes
Test of significance: HC versus CLL (cholesterol), p = 0.05 ' ; HC versus
CLL (C/PL), p
(nmol/mg)681
(nmol/mg)451
±0.09 (4)''
148C369
±
±176
HC
83370
±
204681
636 ±
0.60 ±0.10
(4)0.55
Monocyte
Lymphocytes
Normal
± 57
±160
±0.06 (7)
599 ±198C/PL0.66
0.54 ±0.09(11)
CLLCholesterol 317 ± 88Phospholipid
Student's f test.
b Mann-Whitney U test using a one-tailed critical test.
0 Mean ±S.D.
d Numbers in parentheses, number of determinations.
had the highest P values among the cell types examined (Table
4). The lowest P values were found with CLL plasma mem
branes, while intermediate values were found with normal lym
phocytes and monocytes. These data are consistent with the
whole-cell C/PL ratio results.
The fluorescent polarization of DPH was determined in puri
fied plasma membranes over a temperature range of 10-40°.
The results of these experiments are shown in Charts 1ßand
2 and in Table 5. As has been reported previously for other
cell types, higher polarization values were found for HC mem
branes than for the whole cells from which they originated (8,
25, 34, 35, 44, 48, 49). The P values for HC membranes were
significantly higher than those of CLL membranes (Table 5).
Considerable overlap was seen among the P values obtained
on plasma membranes prepared from normal lymphocytes,
monocytes, and HC.
DISCUSSION
The first point to emerge from the data presented above is
that HC differ from normal or CLL lymphocytes on the basis of
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
VOL. 41
Lipids and Fluidity of HC, Monocytes,
Table 5
Mononuclear cell membrane DPH fluorescence polarization values
Test of significance: CLL versus HC, NL,a MONO, p < 0.001°; CLL versus
NL, MONO, p < 0.01°; CLL versus HC, p - 0.05e.
1.00
o
g
and Lymphocytes
080
o:
2: 0.60
HCMonocyte
O
±0.020
0.298
0.0120.295
±
±0.01 7
(3)e
0.257
(5)0.250±0.020
Lymphocytes
±0.01 9 (7)
±0.017
±0.013
Normal
0.200 ±0.020(10)
0.247 ±0.02037°0.266
CLL15°0.339 0.292 ±0.02325°0.305
1 NL, normal lymphocytes; MONO, monocytes.
0 Student's f test.
c Mann-Whitney U test.
" Mean ±S.D.
e Numbers in parentheses, number of determinations.
040
5
|
±0.020a
0.337
0.0140.329
±
030
i«
QJ
%$°&
¿ 020
HC
MONO
MEMBRANE
NL
CLL
TYPE
Chart 1. Comparison of individual values of mononuclear cell membrane C/
PL ratio and DPH fluorescence polarization at 25°.HC, monocyte (Mono), normal
lymphocyte (NL), and CLL membranes prepared as described in "Materials and
Methods" were analyzed for their C/PL content (Al. Membranes were incubated
with DPH as described in "Materials and Methods," and the fluorescence
polarization at 25" was determined (B).
, mean values of individual determi
nations from each donor for each membrane type. For the case of HC membranes,
multiple determinations were carried out over a time course ranging from several
months to 1 year. Vertical bars, S.D.
Table 4
Mononuclear cell DPH fluorescence polarization values
The results shown represent values averaged from 3 individual CLL lympho
cyte preparations and 2 preparations from the other cell types.
25°
37°
HC
Monocyte
Lymphocytes
Normal
CLL
0.302
0.295
0.251
0.240
0.294
0.259
0.240
0.208
040-
035
0.30
025
EZ! 020
015
15
20
25
30
35
TEMPERATURE (°C)
40
Chart 2. Mononuclear cell membrane DPH fluorescence polarization at 3
temperatures. Membranes from HC, monocytes, normal lymphocytes, and CLL
lymphocytes were prepared and labeled with DPH as described in "Materials
and Methods." A, HC; •¿,
monocyte; O, normal lymphocyte; A, CLL lymphocyte.
Mean values for each membrane preparation at the given temperatures were
obtained either from the average of a minimum of 6 determinations of the DPH
fluorescence polarization at the given temperature or from an extrapolation from
a plot of fluorescence polarization over a temperature range of 10-40°. Bars,
S.D.
OCTOBER
1981
cholesterol content and elevated C/PL ratio. While the similar
ity in the C/PL ratio between HC and normal monocytes would
suggest a monocytoid rather than a lymphoid nature for these
cells, other work from this laboratory has documented charge
differences between HC and normal monocytes following treat
ment with neuraminidase (36). In addition, a recent study by
Golomb et al. (16) has found the protein patterns for HC to be
distinct from those of CLL, acute lymphocytic leukemia, and
poorly differentiated
lymphocytic
lymphoma lymphocytes.
Taken together, the results support a uniqueness for the HC
but do not rule out the strong possibility suggested by a
majority of the published data (reviewed Ref. 20) that the HC
may be a precursor or altered lymphoid cell. It is unfortunately
impossible to compare the properties of the HC with those of
its still undetermined cell of origin.
HC have been shown to display greater than normal cap
formation which is independent of energy production (7, 45).
Since capping depends on the cross-linking of surface recep
tors and involves intracellular contractile proteins, the HC mem
brane may facilitate interaction of the surface receptor with the
corresponding cytoplasmic cytoskeletal components. This con
cept is supported by the finding that enrichment of rabbit
splenic lymphocytes with cholesterol increases their cap for
mation (32). Conceivably, the HC lipid composition may influ
ence its tissue localization in vivo. The cholesterol-rich
HC
membranes may impair the ability of the cells to traverse the
splenic sinusoids and account for the abundance of HC in this
organ (4). Furthermore, a preliminary report by Dawson and
Golomb (12) has described high levels of arachidonate and low
levels of linoleate and oleate which also may affect membrane
fluidity. As such, more data on the fatty acid composition and
types of phospholipids present in the HC must be obtained
before more definite statements can be made.
The data reported above may be relevant to Inbar's concept
that neoplastic lymphoid cells show an increased membrane
fluidity, a theory which was evolved over the last 5 years in a
number of laboratory studies (8, 35, 48, 50). Other investiga
tors have suggested that different cancers may have a discrete
range of whole-cell DPH P values related to the degree of
differentiation of the leukemic cell (44, 49). While we clearly
recognize the uncertainty as to where this probe really is within
the lipid-protein membrane microenvironment (31 ), our findings
with HC give the first example of a human leukemic population
found to possess membranes more rigid than those of CLL
lymphocytes. The observation that HC, which fulfill other cri
teria of neoplasia, have a membrane fluidity equal to or greater
than normal cells indicates that increased fluidity is not a sine
qua non for malignant change.
4053
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
L F. Liebes et al.
Additional data generated by the present study concern the
properties of normal and CLL lymphocytes. The C/PL ratio of
0.44 ±0.07 for whole-cell CLL lymphocytes is in close agree
ment with the value of 0.39 reported by Gottfried (17), rather
than the figure of 0.6 reported by Pratt ef al. (37). The choles
terol content and C/PL ratio of normal lymphocytes in the
present study was lower than the value of 0.71 previously
reported by Gottfried (17). This probably reflects the more
rigorous removal of platelets and monocytes from the postFicoll-Hypaque mononuclear preparation by current method
ology. This interpretation is supported by our finding that
identical values were obtained from our standard preparation
of heparinized blood from which platelets were removed by
multiple washes with low-speed centrifugations of the FicollHypaque mononuclear cells and from defibrinated blood
where, in addition to the above steps, coagulation alone had
removed >90% of platelets in the first step.
When plasma membrane preparations were compared, C/
PL ratio differences between normal and CLL whole-cell prep
arations were no longer apparent. These results may indicate
that cytoplasmic rather than plasma membrane lipids are re
sponsible for the difference between normal and CLL lympho
cytes. In a similar manner, Spiegel ef al. (44) found that the
differences in P values between lymphoblasts of diverse origin
and normal lymphocytes disappeared when membrane prepa
rations were made. Johnson and Robinson (26) also found no
difference between either the lipid composition or the polari
zation values of CLL and normal tonsil membranes and lipo
somes prepared from these membranes.
The lower than normal fluorescence polarization values ob
tained with whole CLL lymphocytes are in reasonable agree
ment with the P values reported by Inbar ef al. (21) (0.277 ±
0.005, 25°), by Spiegel ef al. (44) (0.269 ±0.012, 25°),and
by Johnson and Kramers (25) (0.224 ±0.008, 37°)for CLL
lymphocytes.
The membrane fluorescence polarization values for normal
lymphocyte plasma membranes, obtained in the present stud
ies, overlap the P values reported at 37° by Johnson and
Kramers (0.244 ±0.009) and were somewhat lower than the
value of 0.321 reported at 25° by Petitou ef al. (25, 35). A
wide range with a significantly lower than normal P value was
noted with CLL lymphocyte membranes at all temperatures. In
this respect, it must be remembered that P values also reflect
the degree of unsaturation and length of fatty acids, the classes
of neutral lipids and phospholipids, and the degree of lipidprotein association of intrinsic membrane proteins.
Several alterations in the lipids of CLL lymphocytes have
been demonstrated. Gottfried (18) reported lower levels of
sphingomyelin and lysolecithin for CLL lymphocytes when com
pared with normal lymphocytes. Sphingomyelin has been sug
gested as a factor in restricting membrane fluidity by Inbar ef
al. (21 ) and by Cullis and Hope (11 ). Dawson and Golomb (12)
reported in a comparison with their HC data that CLL lympho
cytes had elevated amounts of oleate and linoleate with rela
tively low levels of arachidonate. A lower level of triglycérides
and a higher level of free fatty acids were found in CLL than in
tonsil lymphocytes (26).
Lipid-protein interactions may also affect fluorescence po
larization. The effect of proteins on membrane fluidity has been
elucidated from model studies with the transmembrane-spanning polypeptide gramicidin A (5, 10). A recent study has
4054
shown this protein to exert a condensing effect on membrane
lipids similar to that observed with cholesterol (47). Through
the use of freeze fracture analysis of normal and CLL lympho
cytes, we have found differences in the behavior of intramembranous particles (52) in that the intramembranous particles of
CLL lymphocytes were much less susceptible to displacement
than were the intramembranous particles of normal lympho
cytes. Conceivably, these anomalous structural-dynamic prop
erties of CLL lymphocytes may be related to the altered lipidprotein interrelationships described above.
ACKNOWLEDGMENTS
The authors wish to thank Drs. K. Rai and A. Sawitzky for providing blood
samples from patients with HCL and Drs. S. Ip and G. Witz for kindly allowing us
the use of their instruments.
APPENDIX
The following instrumental conditions were used in the Hitachi Perkin-Elmer
MPF3: (a) the excitation and emission monochrometers were fixed at 360 and
425 nm, respectively; (t>) fluorescence polarization was monitored through the
respective use of a Polacoat and HNBP polarization filter in the excitation and
emission analyzers. Based on fluorescence intensity measurements, excitation
slits were set at 3 to 5 nm, and emission slits were set at 8 to 10 nm; (c) 5-sec
intervals were used for measurements of vertical and horizontal polarized com
ponents of light followed by 50 sec during which the shutter remained closed.
The Elscint MV-1 microviscometer simultaneously analyzes vertically (/o o) and
horizontally (/ . ) polarized emission intensities (/) using a vertically polarized
excitation beam and displays the degree of steady-state fluorescence polarization
anisotropy according to the relation:
_ _ 'o.o ~ '0,90
'o.O
+
'o.90
The steady-state fluorescence polarization anisotropy of labeled membranes
using the Hitachi Perkin-Elmer was calculated by the expression:
'o.O
—¿G/0.90
'o.o + G'o 00
where G is a grating correction factor that was determined for each instrumental
setting, is equal to 'go.o/'goso and is obtained with the excitation polarization in a
horizontal direction. The variation in P values for a given sample over multiple
determinations taken over 16 months was ±0.006 to 0.009 P.
Data obtained with the Hitachi Perkin-Elmer MPF3 were corrected for the
transmission characteristics of the polarizer filters at the above given excitation
and emission wavelengths to allow comparison with the Elscint MV-1 data.
Polarization measurements on the same sample obtained with both instruments
remained within the variation found using a single instrument.
REFERENCES
1. Allain, C. C., Poon, L. S., Chan, C. S. G., Richmond, W., and Fu, P. C.
Enzymatic determination of total serum cholesterol. Clin. Chem., 20: 470475, 1974.
2. Bohlen, P., Stein, S., Dairman, W., and Udenfiend, S. Fluorometric assay of
proteins in the nanogram range. Arch. Biochem. Biophys.. 155: 213-220,
1973.
3. Braylan, R. C.. Jaffee, E. S.. Triche. T. J., Nanba. K., Fowlkes, B. J.,
Metzger, H., Frank. M. M.. Dolan. M. S., Yee. C. L.. Green, I., and Berard,
C. W. Structural and functional properties of the hairy cells of leukemic
reticuloendotheliosis.
Cancer (Phila.), 41: 210-227, 1978.
4. Burke, J. S., Byrne, G. E.. Jr., and Rappaport, H. Hairy cell leukemia
(leukemic reticuloendotheliosis).
I. A clinical pathologic study of 21 patients.
Cancer (Phila.), 33 1399-1410,
1974.
5. Chapman, D., Cornell, B. A., Eliasz. A. W., and Perry, A. Interactions of
helical polypeptide segments which span the hydrocarbon region of lipid
bilayers. Studies of the gramicidin A lipid-water system. J. Mol. Biol., 113:
517-538, 1977.
6. Chen. P. S., Toribara. T. Y., and Warner, H. Microdetermination of phospho
rus. Anal. Chem., 28. 1756-1758.
1950.
7. Cohen. H. J., George, E. R., and Kremer. W. B. Hairy cell leukemia: cellular
characteristics
including surface immunoglobulin dynamics and biosyn
thesis. Blood, 53. 764-775, 1979.
8. Collard, J. G., DeWildt. A., Oomen-Meulemans. E. P. M., Smeekens, J.,
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
VOL.
41
Lipids and Fluidity of HC, Monocytes,
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Emmelot, P., and Inbar, M. Increase in fluidity of membrane lipids in lym
phocytes, fibroblasts and liver cells stimulated for growth. FEBS Lett., 77:
173-178, 1977.
Cooper. R. A. Abnormalities of red cell membrane lipids. In: R. Silber, J.
LoBue, and A. G. Gordon (eds.), The Year in Hematology, pp. 69-101.
Plenum Publishing Corp., New York: 1978.
Cornell, B. A., Sacre, M. M., Peel, W. E., and Chapman, D. The modulation
of lipid bilayer fluidity by intrinsic polypeptides and proteins. FEBS Lett., 90:
29-35, 1978.
Cullis, P. R., and Hope, M. J. The bilayer stabilizing role of sphingomyelin in
the presence of cholesterol. A 31PNMR study. Biochim. Biophys. Acta, 597.
533-542, 1980.
Dawson, G., and Golomb, H. M. Fatty acid differences amongst leukemia
cell types, with special reference to hairy cell leukemia. A preliminary report.
Clin. Chim. Acta, 96. 85-90, 1979.
Debusscher, L., Bernheim, J. L., Collard-Ronge, E., Govaerts. A., Hooghe,
R., Lejeune, F. J., Zeicher, M., and Strychmans, P. A. Hairy cell leukemia:
functional immunologie, kinetic and ultrastructural characterization. Blood,
46:495-507,
1975.
Folch, J., Lees, M., and Sloan-Stanley, G. H. A simple method for the
isolation and purification of total lipids from animal tissues. J. Biol. Chem.,
226 497-509, 1957.
Golde, D. W., Stevens, R. H., Ouan, S. G., and Saxon, A. Immunoglobulin
synthesis in hairy cell leukemia. Br. J. Haematol., 35: 359-365. 1977.
Golomb, H. M., Nathans, A. H., and Dawson, G. Comparison between the
cellular proteins of hairy cell leukemia and the leukemic phase of other
lymphoproliferative diseases. Cancer Res., 40: 1907-1914,
1980.
Gottfried, E. L. Lipids of human leukocytes: relation to cell type. J. Lipid
Res., 8: 321-327, 1967.
Gottfried, E. L. Lipid patterns in human leukocytes maintained in long-term
culture. J. Lipid Res., 12: 531-537, 1971.
Heider. J. G., and Boyett, R. L. The picomole determination of free and total
cholesterol in cells in culture. J. Lipid Res., 19: 514-518. 1978.
Hooper, W. C., Buss, D. H., and Parker, C. L. Leukemic reticuloendotheliosis
(hairy cell leukemia): a review of the evidence concerning the immunology
and origin of the cell. Leukemia Res., A: 489-503, 1980.
Inbar. M., Goldman, R.. Inbar, L.. Bursuker, I., Goldman, B.. Akstein. E.,
Segal, P., Ip, S . and Ben-Bassat, I. Fluidity difference of membrane lipids
in human normal and leukemic lymphocytes as controlled by serum com
ponents. Cancer Res., 37: 3037-3041,
1977.
Jaffe, E. S., Shevach. E. M., Frank, M. M., and Green, I. Leukemic reticu
loendotheliosis: presence of a receptor for cytophilic antibody. Am. J. Med..
57. 108-114, 1974.
Jansen, J., Schuit. H. R. E., van Zwet, T. L., Meijer, C. J. L. M., and Hijmans.
W. Hairy cell leukemia: A B-lymphocytic disorder. Br. J. Haematol., 42: 21 33, 1979.
Johnson, S. M. A new specific cholesterol assay gives reduced cholesterol/
phospholipid molar ratios in cell membranes. Anal. Biochem., 95. 344-350,
1979.
Johnson, S. M., and Kramers, M. Membrane microviscosity differences in
normal and leukemic human lymphocytes. Biochem. Biophys. Res. Com
mun., 80: 451-457, 1978.
Johnson, S. M., and Robinson, R. The composition and fluidity of normal
and leukemic or lymphomatous lymphocyte plasma membranes in mouse
and man. Biochim. Biophys. Acta, 558. 282-295, 1979.
King, G. W.. Hurtubise. P. E., Sagone, A. L., LoBuglio, A. F., and Metz, E.
N. Leukemic reticuloendotheliosis:
a study of the origin of the malignant cell.
Am. J. Med., 59. 411-416, 1975.
Lopes, J., Nachbar, M., Zucker-Franklin,
D., and Silber. R. Lymphocyte
plasma membranes: analysis of proteins and glycoproteins by SDS-gel
electrophoresis. Blood, 41: 131-140, 1973.
Lopes, J.. Zucker-Franklin, D., and Silber. R. Heterogeneity of 5' nucleotid-
ase activity in lymphocytes in chronic lymphocytic leukemia. J. Clin. Invest.,
52: 1297-1300,
1973.
30. Lowry, O. H., Rosenbrough, N. J., Farr, A. L., and Randall, R. J. Protein
measurement with the Folin phenol reagent. J. Biol. Chem., 793: 265-275,
1951.
31. Mely-Goubert, B., and Freedman, M. H. Lipid fluidity and membrane protein
monitoring using 1,6-diphenyl 1,3,5-hexatriene.
Biochim. Biophys. Acta.
OCTOBER
1981
and Lymphocytes
607 315-327, 1980.
32. Ostro, M. J., Bessinger, B., Summers, J. F.. and Dray. S. Liposome modu
lation of surface immunoglobulins on rabbit spleen cells. J. Immunol., 124:
2956-2965.
1980.
33. Palutke, M.. Weise, R. W., Tabaczka, P.. Varadachari, C., and Axelrod, A.
Hairy cells and macrophages: a comparative study. Lab. Invest., 39. 267280, 1978.
34. Pessin, J. E., Salter, D. W., and Glaser, M. Use of a fluorescent probe to
compare the plasma membrane properties in normal and transformed cells.
Evaluation of the interference by triacylglycerols
and alkyldiacylglycerols.
Biochemistry, 17: 1997-2004.
1978.
35. Petitou, M., Tuy, F.. Rosenfeld. C., Mishal, F., Paintrand, M., Mathe, G.,
Jasnin, C., and Inbar, M. Decreased microviscosity of membrane lipids in
leukemic cells: two possible mechanisms. Proc. Nati. Acad. Sei. U. S. A.,
75: 2306-2310,
1978.
36. Petty, H. R., Liebes, L. F.. Pelle, E., Ware. B. R., and Silber, R. Electrophoretic mobility distributions of hairy cells and chronic lymphocytic leukemia
lymphocytes: a comparison with normal lymphocytes and monocytes. Blood,
57:250-255,
1981.
37. Pratt, H. P. M., Saxon. A., and Graham, M. L. Membrane lipid changes
associated with malignant transformation and normal maturation of human
lymphocytes. Leukemia Res., 2: 1-10, 1978.
38. Rabinovitch, M., and DeStefano, M. J. Use of the local anesthetic lidocaine
for cell harvesting and subcultivation. In Vitro (Rockville). JJ: 379-381,
1975.
39. Rudel, L. L., and Morris, M. D. Determination of cholesterol using o-phthalaldehyde. J. Lipid Res., 14: 364-366, 1973.
40. Scheinberg, M., Brenner, A. I., Sullivan. A. L.. Cathcart, E. S., and Katayama,
I. The heterogeneity of leukemic reticuloendotheliosis.
"hairy cell leukemia."
Evidence for its monocytic origin. Cancer (Phila.), 37. 1302-1307.
1976.
41. Schrek, R.. and Donnely. W. J. Hairy cells in blood in lymphoreticular
neoplastic disease and "flagellated"
cells of normal lymph nodes. Blood.
27: 199-211, 1966.
42. Segel, G. B., Lichtman. M. A., Gordon, B. R., MacPherson, J. L., and
Nussbacher, J. Plateletpheresis residues: a source of large quantities of
human blood lymphocytes. Transfusion, 16: 455-459,
1976.
43. Shevach. E. M., Herberman, R., Frank, M. M.. and Green, I. Receptors for
complement and immunoglobulins on human leukemic cells and human
lymphoblastoid cell lines. J. Clin. Invest.. 5Õ: 1933-1938.
1972.
44. Spiegel, R. J., Magrath, I. T.. and Shutta, J. A. Role of cytoplasmic lipids in
altering diphenylhexatriene
fluorescence polarization in malignant cells.
Cancer Res., 41. 452-458. 1981.
45. Splinter, T.A. W., Collard, J. G., DeWildt, A., Temmink, J. H. M.. and Decary,
F. Capping of surface immunoglobulin on "hairy cells" is independent of
energy production. J. Cell. Sci., 36: 45-59, 1979.
46. Stein. H., and Kaiserling, E. Surface ¡mmunoglobulins and lymphocytespecific surface antigens on leukemic reticuloendotheliosis
cells. Clin. Exp.
Immunol., 78:63-71,
1974.
47. Susi, H., Sampugna, J.. Hampson, J. W.. and Ard, J. S. Laser-Raman
investigation of phospholipid-polypeptide
interactions in model membranes.
Biochemistry. 18: 297-301. 1979.
48. Van Blitterswijk. W. J., Emmelot, P., Hilkmann. H. A. M.. Oomen-Meulemans,
E. P. M., and Inbar, M. Differences in lipid fluidity among isolated plasma
membranes of normal and leukemic lymphocytes and membranes exfoliated
from their cell surface. Biochim. Biophys. Acta. 467: 309-320. 1977.
49. Van Hoeven, R. P.. Van Blitterswijk. W. J., and Emmelot, P. Fluorescence
polarization measurements on normal and tumor cells and their correspond
ing plasma membranes. Biochim. Biophys. Acta. 55Õ: 44-54, 1979.
50. Yanovich, S., Harris, K., Sallan, S. E., Schlossman, S. F., and Inbar, M.
Dynamic parameters of membrane lipids in normal and leukemic human
lymphocytes isolated from peripheral blood and bone marrow. Cancer Res.,
38:4654-4661,
1978.
51. Zucker-Franklin,
D. The percentage of monocytes among "mononuclear"
cell fractions obtained from normal human blood. J. Immunol., 7 72. 234240. 1974.
52. Zucker-Franklin, D., Liebes, L. F., and Silber, R. Differences in the behavior
of the membrane and membrane-associated filamentous structures in normal
and chronic lymphocytic leukemia (CLL) lymphocytes. J. Immunol., 722:
97-107, 1979.
4055
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
L F. Liebes et al.
Fig. 2. A, membrane preparation of normal monocytes; ß,membrane preparation of HC. x 55,000.
4056
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.
VOL. 41
Comparison of Lipid Composition and
1,6-Diphenyl-1,3,5-hexatriene Fluorescence Polarization
Measurements of Hairy Cells with Monocytes and Lymphocytes
from Normal Subjects and Patients with Chronic Lymphocytic
Leukemia
Leonard F. Liebes, Edward Pelle, Dorothea Zucker-Franklin, et al.
Cancer Res 1981;41:4050-4056.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/41/10/4050
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1981 American Association for Cancer Research.