1382 Blood. Vol 71, No 5 (May), 1988: pp 1382-1389
Reed-Sternberg Cells in Hodgkin’s Cell Lines HDLM, L-428, and KM-H2 Are notActively Replicating: Lack of Bromodeoxyuridine Uptake
by Multinuclear Cells in Culture
By Su-Ming Hsu, Xun Zhao, Subendu Chakraborty, Y-Fa Liu, Jacqueline Whang-Peng, Ming S. Lok, and Shiroh Fukuhara
We compared the proliferation of mononuclear and multi-
nuclear cells in four Hodgkin’s cell lines, HDLM-1 , HDLM-
1 d, 1-428, and KM-H2, by examining their capacity to
incorporate bromodeoxyuridine (BrdUrd) into nuclei.
Approximately 5% of all cells in HDLM-1 cultures had two
or more nuclei, a characteristic of Reed-Sternberg (RS)
cells. Unlike mononuclear Hodgkin’s (H) cells, these RS
cells exhibited no uptake, or only minimal uptake of
BrdUrd, suggesting that they did not replicate actively.
Cytogenetic study showed that 25% of the HDLM-1 cells
contained a tetraploid (4X) set of chromosomes with a
characteristic two-peak distribution. Following treatment
of HDLM-1 cells with phorbol ester. the percentages of 4X
cells and RS cells increased to 50% and 12%. respectively.
This increase in RS cells was not likely to be due to cell
fusion as shown by the absence of hybridization of BrdUrd-
positive and -negative nuclei. Phorbol ester has a short-
term effect of blocking the exit of cells from G1 into S
phase, but no effect on the transition from S phase to G2IM
phase. The block is more prominent in 2X cells than in 4X
cells, which may explain the increase in percentage of 4X
T HE MONONUCLEAR Hodgkin’s (H) cells and the
binuclear or multinuclear Reed-Sternberg (RS) cells
display similar, if not identical, markers.’5 These two types
of cells make up the neoplastic cells in Hodgkin’s disease
(HD). However, the mechanism involving the transforma-
tion of H cells to RS cells is not yet completely known. Kadin
and Asbury6 have attributed the formation of RS cells to
abnormal multipolar mitoses and disturbed cytokinesis
rather than to cell fusion.
In earlier investigations, RS cells were thought to be
end-stage cells that, although important for the diagnosis of
HD, had little or no capacity to replicate.78 This conclusion
was based on the fact that mitoses and tritiated-thymidine
uptake could be seen in H cells, but only rarely, if at all, in
RS cells. Later autoradiographic studies on long-term cul-
tunes of Hodgkin’s tissue indicated that RS cells are capable
of DNA synthesis.6’9 However, these studies, in which cul-
From the Department of Pathology. University of Texas Health
Science Center at Houston; Cytogenetic Oncology Section. Mcdi-
cine Branch, National Cancer Institute, Bethesda, MD; Hematolo-
gy-Oncology Service, Veterans Administration Hospital, Leaven-
worth, KS; and First Division of Internal Medicine, Faculty of
Medicine, Kyoto University. Japan.
Submitted July 20, 1987; accepted January 4. 1988.
Address reprint requests to Su-Ming Hsu. MD, Department of
Pathology. University of Texas Health Science Center at Houston.
P0 Box 20708, Houston, TX 77225.
The publication costs ofihis article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. §1 734 solely to
indicate this fact.
C I 988 by Grune & Stratton, Inc.
0006-4971/88/7105-0026$3.OO/O
cells in phorbol ester-treated cultures. In addition, phorbol
ester induced the differentiation of H-RS cells, which was
accompanied by loss of the marker HeFi-1 from the cell
surface. Approximately one third of the RS cells did notexpress HeFi-1 , or expressed only minimal amounts. The
findings led us to the following conclusions: (1 ) The 4X cellsprobably are formed from 2X H cells as a result of disturbed
cytokinesis, but not a cell fusion. (2) A considerable num-
ber of 4X cells were H cells, because the number of 4X cells
consistently exceeded that of RS cells. (3) Since mitotic
figures are extremely rare in RS cells and these cells did
not show active BrdUrd uptake, the increased number of
RS cells must also be a consequence of disturbed cytokine-
sis of H cells or a result of nuclear transformation (twisting,
convolution, or separation of the nucleus) in H cells. (4)
Most RS cells lose their proliferating capacity and some RS
cells may undergo further differentiation. Uptake of BrdUrd
and phorbol ester induction were also studied on the other
three H-RS cell lines, HDLM-1 d. 1-428, and KM-H2, with
results similar to those for HDLM-1.
0 1988 by Grune & Stratton, Inc.
tuned cells prepared from HD tissue were used, may have
suffered from serious artifacts. For example, fibroblasts and
Epstein-Barr virus-transformed B cells tend to overgrow in
culture.’#{176}Although the neoplastic nature ofthe cultured cells
in one study was clearly confirmed by the finding of cellular
aneuploidy and by the capacity of the cells to grow in the
brains of nude mice,9 contaminating cells may have been
present, some of which could display cytologic characteris-
tics resembling those of RS cells. Such artifacts may
preclude a precise evaluation of the number of true RS cells
in culture.
Recently, the establishment of H-RS cell lines and the
development of monoclonal antibodies (MoAbs) specific for
H-RS cells have made studies of the biologic nature of these
cells easier than before. The study of H-RS cells can yield
more meaningful and interpretable results than as obtainable
with heterogeneous cell cultures.
In a previous ‘ we observed that the percentage of
RS cells increased following the I 2-O-tetra-decanoyl phor-
bol- 1 3-acetate (TPA)-induced differentiation of H-RS cells.
Therefore, we became interested in investigating the mecha-
nism of RS cell formation. In the present study, we showed
that the increase in the percentage of RS cells following TPA
induction paralleled the decreased proliferation of the cul-
tuned H-RS cells. This observation raises the question
whether RS cells are capable of replication. We tested the
capacity of H cells and RS cells to take up bromodeoxyurid-
inc (BrdUrd), an analogue of thymidine that is incorporated
into the nuclei of replicating cells. Unlike H cells, RS cells
took up BrdUrd only minimally, indicating such cells usually
do not proliferate. Furthermore, we obtained evidence that
cell fusion is not responsible for the formation of RS cells.
Instead, RS cells could be generated as a result of disturbed
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
LACK OF BRDURD UPTAKE IN RS CELLS 1383
cytokinesis, as has been suggested in previous studies,6 or by
twisting and convoluting of the nucleus.
MATERIALS AND METHODS
Tissue culture ofH-RS cells. HDLM is a series of four cell linesderived from a patient with HD. All ofthesecell lines (HDLM-l, -2,
-ld, -2d) have similar, if not identical, phenotypes and cytogeneticmarkers.’2’5 The HDLM-ld and -2d cells had been cultured in thepresence of phorbol ester (TPA) for more than 1 year; they were then
cultured in a TPA-free medium. We used HDLM-l and HDLM-ldcells in the present study because, in our laboratory, these cells grew
more rapidly than did HDLM-2 and -2d cells. In addition, we used
two H-RS cell lines, L-428 and KM-H2.’#{176}”6These cells show aphenotypic expression (ie, they react with MoAbs Ki-l, IRac, 2H9,
and HeFi-1) similar to that of HDLM cells. All cells were grown at 4
x l0� to 2 x 106 cells/mL in RPM! 1640 medium (GIBCO, GrandIsland, NY) supplemented with 10% fetal calf serum, 2 mmol/Lglutamine, 50 �mol/L 2-mencaptoethanol, and 50 j.tg/mL gentamy-
cm at 37#{176}Cin a humidified, 5% CO2 atmosphere. The medium was
changed every two to three days. The viability of these cells has
generally been maintained at 95%. Cell viability was determined by
the Trypan Blue dye exclusion test.All of these cells have a doubling time of approximately 60 to 84
hours. Smears were prepared by cytocentnifugation during thestationary phase of culture and were stained with Diff-Quik forcytologic evaluation. A total of 4,000 cells were counted, and thenumber of RS cells among them was identified as those cells
containing two or more nuclei that were either widely separated or inintimate contact. The result was expressed as the number of RS cells
per 100 cells in culture.Ultrastructural study of HDLM-1 cells. HDLM- 1 cells were
fixed in 3% glutanaldehyde, embedded in Epon, and processed forelectnon-microscopic examination of the nuclear structure.
MoAbs and immunoperoxidase reagents. To compare the phe-
notypes of mononuclear H cells and RS cells, we used the MoAbsHeFi-l , 2H9, and anti-IRac for immunostaining on cytospin smears.
The specificities of these MoAbs and the staining procedure havebeen described in detail previously.5’72#{176} The labeling reagents usedwere biotin-labeled horse anti-mouse Ig and avidin-biotin-penoxi-dase complex (ABC), both obtained from Vector Laboratories
(Burlingame, CA). In addition, we also examined the expression of anuclear antigen, Ki-67, in these cells. Ki-67 has been reported to bean antigen for proliferating cells.2’
TPA induction of cells. To study the effect of TPA on theformation of RS cells, we used a protocol for TPA induction ofHDLM-1 cells that has been described previously.”22 Briefly, TPA
dissolved in DMSO (14 �.tg/mL) was added at a final concentrationof 2 i�g/mL to cultures of H-RS cells. Every second day, two-thirdsof the medium was replaced with fresh medium containing TPA.The induction was carried out for three to four days. Cytospinsmears were prepared during the course of induction and wereevaluated for the percentage of RS cells. The increase in the numberof RS cells was also determined with night-angle light scatter, whichwe used as a detection parameter in flow cytometry.23
Effect of TPA on cell cycle. To determine the effects of TPA on
cell proliferation, we examined the DNA cycle of HDLM-l andHDLM-ld cells by using propidium-iodide staining and flow cytom-
etry.’5’22 Briefly, the cells were cultured in medium with or without
TPA for two days. The cells were washed in RPM! 1640 medium,
and then 1 mL aliquots of suspensions containing 2 x 106 cells wereincubated with 95% ethanol (2.5 mL) for 24 hours at 2 to 8#{176}C.After
being washed in phosphate buffered saline (PBS), the cells were
incubated with 100 ML RNase A (1,000 U/lOO �tL, WorthingtonDiagnostic, NJ) for one minute. The samples were then stained with
propidium iodide (50 �ig/mL) for two minutes and vortexed gently.Finally, the cells were analyzed by flow cytometny (Ontho Cytofluor-
ograf system 50H). The phase of TPA-tneated cells in the DNA
cycle was compared with that of control cells.BrdUrd uptake. HDLM-I, HDLM-ld, L-428, and KM-H2
cells were cultured in the presence of BndUrd (1 x 106 mol/L) forvarious periods (one to 72 hours). The cells were washed with Tnis
buffered saline (TBS) (0.1 mol/L, pH 7.6), prepared as cytospinsmears, and examined for uptake of BrdUrd by the nuclei. We used
the ABC immunoperoxidase method with an anti-BrdUrd MoAb(Becton Dickinson, Sunnyvale, CA) to detect BrdUrd uptake in the
nuclei.2426 Briefly, smears were treated with 1 mol/L HCI in normal
saline for 30 minutes and then washed with TBS for ten minutes.
Next, the cells were stained with anti-BndUnd MoAb ( 1:40), biotin-labeled horse anti-mouse IgG, and ABC as described previously)”26
Cells that take up BrdUnd exhibit dank granular staining in thenuclei. At various intervals, the percentage of BndUrd uptake by RScells was compared with that by mononuclear cells.
To test for possible fusion of mononuclear cells, we took cellspreviously treated with BrdUrd for 36 hours and cultured them withcontrol, untreated cells for one to two days. The RS cells were thenexamined for the presence of hybrid nuclei (BrdUrd-positive nucleusand Brd-Urd-negative nucleus in the same cell).
Cytogenetic study. We studied the chromosome distribution in
both TPA-treated and control HDLM-l cells. Both TPA-inducedand control cells were treated with colcemide (0. 1 �ig/mL, 90minutes) and hypotonic solution and were examined for the number
of chromosomes that they contained. The colcemide treatment did
not affect the number of RS cells in culture. A total of 2 15 cells wereincluded.
Sorting ofcells with high DNA content. We sorted HDLM-ld
cells according to their DNA content and determined the percent-
ages of RS cells in populations with high and in those with low DNAcontent. The sorting procedure was carried out with an OrthoCytofluorognaf 50H controlled by a 21 50 computer system that wasattached to the machine. Cells with high (highest 25%) and lowDNA content were deflected into separate containers, centrifuged
onto slides, and stained with Diff-Quik. The percentage of RS cellsin each sample was then determined as described in a previoussection (Tissue culture of H-RS cells).
Single-cell culture. The uptake of nucleic acid into cell nucleidoes not necessarily indicate that these cells can successfully com-plete the replication cycle. We used a limiting-dilution technique toculture single cells in 96-well plates to determine the replicationcapacity of the cells. We examined each well under a phase contrastmicroscopy to determine the number of cells and the size of cell.
Cells with multiple nuclei generally had a size three to four timesthat of mononuclear cells. Frequently, the number of nuclei in thelarge cell can be determined by the presence of two or more refractilenucleoli. Wells that contained more than one cell and those in whichthe number of nuclei per cell could not be ascertained were excludedfrom the study. The plates were cultured for ten to 14 days andexamined for the formation of colonies in each well.
RESULTS
Number of RS cells in culture and their morphologic
characteristics. The percentages of RS cells varied among
the four cell lines studied. From 7% to 1 1% of HDLM-ld and
L-428 cells were RS cells, followed by 4% to 6% for HDLM-
1, and 3% to 5% for KM-H2 cells, respectively (Fig 1). The
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
EJD0
30
UCo
0
Control TPA
Fig 1 . Percentages of RS cells in control and TPA-treatedH-RS cells in cultures. The result obtained with 1-428 is similar tothat of HDLM-ld.
percentage of RS cells remained constant in control cultures
throughout the 18-month study period.
Most of the RS cells contained two nuclei that were eitherwidely separated or in contact. Often, the two nuclei were
connected to one another by a thin, thread-like substance. A
highly twisted nucleus was frequently observed in H cells;
these cells presumably became RS cells after going through
several stages of transformation (Fig 2). The twisting and
deep convolution of nuclei were apparent on electron micros-
copy (Fig 3).
Phenotype ofRS cells v H cells. The phenotypic expres-
sion of RS cells was similar, if not identical, to that of H cells.
Both expressed HeFi-l, 2H9, and IRac. The staining inten-
sity of the H cells was fairly uniform; however, the expression
of HeFi-l in RS cells was quite variable. The intensity of
HeFi- 1 staining in 20% to 35% of the RS cells was very weak
or negative (Fig 4), but, in 10% to 20% of these cells, itappeared to be more intense than that in H cells. The
TPA-induced RS cells and H cells both had diminishedexpression of HeFi-l and 2H9, but remained IRac-positive.
The expression of Ki-67 was highly variable among cell
lines (Table 1 ). In L-428, most H as well as RS cells showed
� �
�I �#{248}
A, W ,�. E”W’,� ‘�
Fig 2. Nuclei of HDLM-1 cells vary morphologically. (A-C) Thenuclei appear to be twisted, but remain connected. (D-F) the two
nuclei began to separate. but remain attached to each other by athin thread. (G) the connection has been broken. Original magnifl-
cation x 400.
Fig 3. Electron microscopy demonstrates the highly indented
or convoluted nuclei in HDLM-1 cells. The two nuclei in a cell may
appear separate or connected to one another depending on theplane of the section.
variable staining. In HDLM, intense staining was observed
in most H cells, but a considerable number of RS cells were
not stained on stained only weakly (Fig 5). The staining of
KM-H2 cells was uniformly weak. Treatment with TPA
reduced the staining intensity in all of the H-RS cell lines.
Effect of TPA on the formation of RS cells and on their
phase in the cell cycle. Induction with TPA had a dramatic
effect on the formation of RS cells. In all cultures, the
percentage of RS cells increased during a three-day induc-
tion period (Figs 1 and 6). For example, the percentage of RScells in HDLM-l cultures increased from 4% to 6% to I 1% to
15%. The cell viabilities were slightly lower in TPA-treated
cultures than in control cultures.
The TPA induction also affected the cell replication cycle.The precise percentage of cells in each phase is difficult to
determine because the aneuploidy of chromosome. However,
we estimated that, in control cultures, approximately 25% of
HDLM-l (2X) cells were in GO/GI phase (region 1, Fig 7)
and 20% in S phase (region 2). In TPA-treated cultures, 46%
,
�1#{149}\
�
/,� Lt�i�.
(
0
Fig 4. HeFi-1 staining of HDLM-1 cells. Most H cells expressedthe HeFi-1 antigen. However, approximately 30% of RS cells(arrows) did not express, or expressed only minimal amounts of,
HeFi-1. (A, HDLM-1; B, HDLM-ld; C, KM-H2).
1384 HSU El AL
�#{149}‘:,
‘.1
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
LACK OF BROURD UPTAKE IN RS CELLS 1385
Table 1 . Ki-67 Expression and BrdUrd Uptake in H-RS Cells
HDLM-1 L428 KM-H2
H RSH RS H RS
Ki-67
BrdUrd
>95% (3+)
>90% (3+)30-40% (2+13+) 60-70% (2+13+) 40-50% (2+)
<30% (±) >90% (3+) <30% (±)>90% (+)
>90% (3+)80% (±)
<25% (+)
The stainin 9 was performed on cells in stationary phase (Ki-67) or on cells incubated with BrdUrd for th rae days. 3 + . Strongl y positive; 2 +.
moderately positive; + , weakly positive; - negative.
(region 1) were in GO/GI phase and 6% in S phase. There was
no significant change in regions 4 and 5 (S and G2/M phase
of 4X cells, respectively). These data were interpreted as
showing that TPA enhanced entry of cells into , phase, but
also blocked their exit from late G, into S phase, and the
blocking is more prominent for 2X cells than 4X cells. A
similar GO/GI block was obtained for the other three cell
lines (L-428, HDLM-ld, and KM-H2).
BrdUrd uptake in H-RS cells. The uptake of BrdUrd by
the nuclei of H cells depended on the length of time of
exposure to BrdUrd. Most H cells (85% to 90%) had BrdUrd
in their nuclei after 18 to 24 hours of incubation (Fig 8).
However, the nucleic-acid incorporation in RS cells was
minimal. At 30 to 36 hours, only approximately 25% of RS
cells showed weak, granular or rare scattered staining in
their nuclei (Fig 9 and Table I). Intense BrdUrd uptake was
detected in <5% of RS cells (at 36 hours), and in 25% of the
RS cells after prolonged (>72 hours) incubation with
BrdUrd. The BrdUrd uptake by the cells appeared to be
affected by TPA treatment. Staining with anti-BrdUrd was
weaker in TPA-treated cells than in control cells (not
shown). The results were similar for all four cell lines tested.
In studying the possibility ofcell fusion, we did not observe
any RS cells that contained hybrid nuclei or nuclei with
intermediate staining intensity. Both nuclei in an RS cell
either were not stained or were stained weakly with anti-
BndUrd, indicating that the formation of RS cells was not a
result ofcell fusion.
Cytogenetic study. The numbers of chromosomes in
HDLM-l cells are illustrated in Table 2 and in the histogram
shown in Fig 10. The distribution remained similar through
the 18-month study period. Most HDLM-l cells had chro-
mosome numbers between 30 and 42 (hypodiploid) with
modal numbers at 2 peaks, at 37 to 38 and 74 to 76
chromosomes. Approximately 25% of the cells had 74 on
more chromosomes (tetraploid, 4X). The chromosome distri-
bution was similar in HDLM-l and HDLM-ld cells, even
Fig 5. Ki-67 staining of HDLM-1 cells. Thenuclei of most (>90%) H cells expressed Ki-67,compared with only 30% of the nuclei of RS cells(arrow heads). Few small binucleated RS cellswere positively stained by Ki-67 (arrows).
though the HDLM-ld cells had been treated continuously
with TPA for more than 1 year. TPA treatment apparently
increased the numbers of 4X cells (50%) over those in
untreated (control) cultures (25%).
Correlation ofDNA content with number ofnuclei. We
sorted the cells with the highest DNA content (highest 25%)
and compared the number of nuclei in these cells with that of
the remaining cells, which had a low DNA content. The
percentage of RS cells (7% to 10%) in the high-DNA-
content population was very similar to that for low DNA
content and that for presorted cell samples. This result
indicates that cells with high DNA content (4X cells ?) did
not necessarily contain multiple nuclei.
Capacity ofRS cells to divide. A total of 563 single-cell
cultures were evaluated for their capacity to proliferate.
Among 476 cultured small cells (H cells), 125 (25%) multi-
plied successfully. However, among 87 cultured large cells
(RS cells), only five cells multiplied. It appeared that most
large RS cells with widely separated nuclei did not prolifer-
ate in single-cell culture.
DISCUSSION
By using H-RS cell lines, we have shown that RS cells do
not participate actively in cell replication, and that these are
likely to be end-stage cells, as had been suggested by
Peckham and Cooper.7’8 This is confirmed by the lack of
BrdUrd uptake in the majority of RS cells during three days
of incubation, and by the lack of proliferation of large RS
cells in single-cell culture. Furthermore, we have provided
evidence that cell fusion is not responsible for the formation
of RS cells by showing that no cells containing hybrid
BrdUrd-positive and -negative nuclei were present. Dis-
turbed cytokinesis may contribute to the formation of 4X
cells, and probably of RS cells as well. Other possible
mechanisms of RS cell formation include convolution, twist-
ing, and separation of nuclei.
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
#{163}Mononuclear Cell
0 RS Cell
18
II� t#{149}#{149}s
31-I 31-I
1386 HSU El AL
a
a
a0a.aa
U0
Hours of BUDR IncubatIon
Fig 8. BrdUrd uptake by HDLM-1 cells. The nuclei of most Hcells contained BrdUrd after 24 hours of incubation, comparedwith only 1 5% of the nuclei of RS cells. The numbers of RS cellscontaining BrdUrd increased as the incubation period increased.
Since RS cells are likely to be derived from H cells, the increase inthe number of BrdUrd-positive AS cells may result from a transfor-mation of BrdUrd-positive H cells into AS cells.
Fig 6. Increase in number of AS cells after TPA induction. Thiseffect of TPA was observed best with HDLM-1 d calls. Theincreased number of AS cells can be measured by the increase onright-angle scatter on flow cytometry (A. control HDLM-1 d cells;B. TPA-treated HDLM-ld cells). as indicated by the shift to theright in B. (C) Some of the TPA-induced HDLM-1 d cells that
contained two or more nuclei (arrows).
We observed BrdUrd uptake in virtually all H cells, but in
only 25% of RS cells, after 36 hours ofculture. To reach 50%
uptake, H cells had to be cultured with BrdUrd for more than
I 2 hours. The extent of nucleic-acid incorporation in H cells
in this study was similar to that in a previous study in which
36.5% of cells were labeled after incubation with 3H-thymid-
inc (1 zCi/mL) for 1 7 to 1 8 hours.9 However, Kadin and
Asbury6 reported a similar degree of uptake (24% to 43%) by
H cells when they used the same amount of 3H-thymidine,
but only one hour of incubation. In both studies,6’9 multinu-
clear (RS ?) cells made up between 1% and 2% of the total
number of cells, and 20.7% to 44% of these were labeled. The
difference in the source of cells and the conditions of culture
could account for the difference in results in their studies and
ours.
In our study, we used BrdUrd, a pyrimidine analogue of
thymidine, to examine the DNA synthesis of the cells,24’26 By
using a MoAb to BrdUrd, one can measure the incorporation
of this thymidine analogue into the DNA of H-RS cells
exposed in vitro, without the need for a radioactive isotope.
The method is very sensitive requiring only l0� or l06
mol/L BrdUrd.2�26 Thus, the lack of uptake by RS cells
cannot be attributed to the sensitivity of the method used.
The uptake of BrdUrd by RS cells, if it occurs, is far less
than that by H cells. The staining in RS cells usually consists
of a few granules and specks, unlike the intense, diffuse
Fig 7. Cell cycles of TPA-treated HDLM-1
cells (A. left) and control cells (B. right). Theprecise percentage of cells in each phase of cellcycle cannot be determined because the aneu-ploidy of cells. The cell cycle can be separatedinto five regions. Aegion 1 corresponds to theG0/G1 of 2X cells; region 2. S phase of 2X cells;region 3. G2 /M phase of 2X cells plus G0/G1phase of 4X cells; region 4, S phase of 4X cells;region 5, G2/M phase of 4X cells. Note thatthere is a significant decrease of percentage of
cells in region 2. but not in region 4 followingTPA treatment. Aegions 4 and 5 consisted ofapproximately 27% of total cells in bothtreated and control cultures.
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
LACK OF BRDURD UPTAKE IN RS CELLS 1387
Fig 9. The extent of BrdUrd uptakein AS cells was minimal compared withthat in H cells. (A) A Diff-Quik-stainedsmear of HDLM-1 cells for cytologicevaluation. (B) Most H cells containabundant BrdUrd, whereas several AScells (arrows) without significant up-
take are seen. The same result wasobserved with KM-H2 cells (C). Thefaint nuclear staining is due to themethyl-green counterstain applied tocontrols (D).
staining seen in most H cells. In limiting-dilution cultures,
we rarely observed any proliferation of RS cells. It is
apparent that, even though we observed BrdUrd uptake in a
few (<25%) RS cells, the probability that these will complete
the cellular replication cycle is small because of their mini-
mal uptake of nucleic acid.
There was a considerable discrepancy between the number
of cells positive for BrdUrd and those positive for Ki-67.
Ki-67 is a nuclear antigen for proliferating cells. It has been
reported that H-RS cells in tissues express Ki-67,2’ but
whether the expression alone can be used to indicate the
replicating capacity of cells is not known. In this study, the
staining by Ki-67 in the H-RS cells used yielded some
confusing results. For example, most (>95%) mononuclear
HDLM-l cells were Ki-67-positive even though at least 25%
of them were in GO/GI phase. In KM-H2 cells, the staining
was uniformly weak and did not vary among cells. A
sucessful cell division is a result of multiple, integrated
biochemical processes, and one cannot deduce the entire
replicating function of cells by detecting a single protein,
especially when the cells are neoplastic and have biochemical
derangements.
Unlike most human neoplastic cell lines, HDLM-l cells
are characterized by a two-peak (2X and 4X) distribution of
the number of chromosomes; this is seen also for other H-RS
cell lines, including L-428, L-439, L-591, and possibly,
KM-H2.’#{176}”6 The precise percentage of 4X cells in culture,
however, cannot be determined. On cytogenetic study, one
can detect only those cells entering the mitotic cycle that are
arrested in metaphase on colcemide treatment. We do not
know whether 2X cells and 4X cells have the same rate of
mitosis. Nevertheless, judging by the DNA content distribu-
tion in our cell cycle analysis, we estimate that the percent-
age of 4X HDLM- 1 cells was in the vicinity of 25%.
Since the number of 4X cells exceeded the number of RS
cells, it is quite clear that a considerable number of 4X cells
have a single nucleus. This conclusion is also supported by
the following observations: (1) The cells with high DNA
content included a large number (90%) of H cells. (2) Most
RS cells had absent or minimal BrdUrd uptake and thus
were not likely to enter the mitotic cycle. It was clear that
most mitotic cells in prophase, which we examined, had only
one nucleus. Since RS cells are not likely to undergo cell
division, it is impossible to determine the number of chromo-
some in RS cells.
The formation of 4X cells is probably attributable to
disturbed cytokinesis of 2X cells, rather than to cell fusion.
The 4X sibling cells may contain two nuclei that are widely
separated (as in some RS cells). The RS cells lose the
capacity to replicate because the two nuclei are incapable of
synchronization in the cell cycle. A large number of 4X cells
may have a single nucleus because of incomplete separation
during mitosis. These cells can undergo mitosis and thus can
be detected on cytogenetic study. Some of the single-
nucleated cell may become RS cells as a result of convolu-
tion, twisting, and subsequent separation of nuclei. We
suspect that mitosis without cytokinesis is a property of some
2X cells, but generally not of 4X cells, because there were
only rare cells with four times (8X) the normal number of
chromosomes (Fig 1 1).
The increases in the percentage of 4X cells in metaphase
and in the percentage of RS cells following TPA induction
are also of interest. TPA can produce a G, block and delay
entry into S phase, whereas cells in S phase can proceed to
G2/M phase. The effect of TPA on the cell cycle of H-RS
cells is similar to that on human peripheral-blood lympho-
cytes, HeLa cells, and the cells of the mouse fibroblast line
C3H/lOTl/2.27’2’ Since the increased number of 4X cells
can be detected as early as two days after TPA induction,
and since the cell doubling time is approximately 60 hours, it
Table 2. Distribution of Chromosome Number in HDLM-1 Cells
<35 36 37 38 39 40 41 42 43-69 70 71 72 73 74 75 76 77 78 79-92 100-150 >200
Total No.of Ce8s
Examined
HDLM-1/TPA 2 3 7 10 1 3 0 0 5 2 1 7 0 8 2 11 3 7 6 5 2 85
HDIM-1 11 1 18 16 8 10 8 3 10 4 3 3 1 3 2 13 0 1 5 8 2 130
Total 13 4 25 26 9 13 8 3 15 6 4 10 1 11 4 24 3 8 11 13 4 215
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
10
0
S
10
0
n �J,n nTPA-treated HDLM- 1 Cells
L1� nfl n n flrJ1�1fJ�l{�ln 1138
Number of chromosomes
Fig 1 1 . Hypothetical scheme of AS-cell formationdue to disturbed cytokinesis and nuclear twisting andseparation. Most AS cells are not likely to enter thecell division cycle. as shown by minimal nucleic-acidincorporation. M + C. mitosis with cytokinesis;M - C. mitosis without cytokinesis; 2X. diploid; 4X.tetraploid.
�i\
Ox
I II� -
1388 HSU El AL
HDLM-1 Cells
a.h� �� flmnJ-.. ..- - . . .. .. 111111#{149}811]
is difficult to attribute the increase to the inhibition of
cytokinesis by TPA, as such an effect would require two cell
cycles to become evident. It appears that the block exerted by
TPA is more prominent on 2X cells than on 4X cells, which
causes a relative increase in the percentage of 4X cells
detected on cytogenetic study.
It is highly unlikely that the 4X and 2X cells represent two
separate clones, because the chromosome distribution is not
much different in HDLM-l and HDLM-ld cells. (HDLM-
ld cells were derived from HDLM-1 cells after incubation of
the latter with TPA for more than 1 year.) Since TPA affects
the mitosis of 4X cells to a lesser degree, one would expect a
gradual replacement of 2X cells by 4X cells in HDLM-ld
culture, if two different clones exist. Phenotypic study by
multiple markers in HDLM-l and HDLM-ld cells has not
shown any significant heterogeneity of cells, or any other
evidence in support of the two-clone theory. Furthermore, a
considerable number of cells in other H-RS cell lines are also
4X. One cannot argue that all H-RS cell lines contain two
distinct populations. Taken together, some of the 4X cells
must be generated continuously from 2X cells as a result of
distributed cytokinesis.
It has been known that TPA induces the differentiation of
H-RS cells.” Most H-RS cells express markers such as
Fig 1 0. Chromosome distribution in untreatedand TPA-treated HDLM-1 cells. Note the character-istic two-peak distribution.
HeFi-l and Ki-l,’7’29 which are considered to be markers of
immature or neoplastic cells because they are absent from
normal lymphoreticular cells.’6 Thus, it is not surprising that,
after TPA induction, the H-RS cells become negative for
these markers.” The increased number of RS cells following
TPA induction may be consistent with the status of cellular
differentiation of H-RS cells, because a small portion of the
RS cells did not express HeFi-l , or expressed it only mini-
mally. However, cellular differentiation cannot totally
explain the mechanism of RS-cell formation, because most of
the RS cells still expressed amounts of HeFi-l antigen
comparable with those of H cells. The formation of RS cells
following TPA induction may result from a decreased rate of
cellular proliferation that allows the twisted nuclei in the
cells to become separated.
In conclusion, the mechanism of RS-cell formation is
probably a multifaceted rather than a simple process. The
disturbed cytokinesis of H cells and the twisting, convolution,
and separation of their nuclei lead to the formation of 4X
cells and RS cells. RS cells do not proliferate to any extent,
even though they take up a small amount of nucleic acid and
express Ki-67 antigen. The capacity of RS cells to complete
the cell division cycle is doubtful.
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
LACK OF BROURD UPTAKE IN AS CELLS 1389
REFERENCES
1. Hsu SM, Yang K, Jaffe ES: Phenotypic expression of Hodg-
kin’s and Reed-Sternberg cells in Hodgkin’s disease. Am J Pathol118:209, 1985
2. Hsu SM, Jaffe ES: Leu Ml and peanut agglutinin stain theneoplastic cells of Hodgkin’s disease. Am J Clin Pathol 82:29, 1984
3. Hsu SM, Zhao X: Expression of interleukin 1 in H-RS cells
and neoplastic cells from true histiocytic lymphomas. Am J Pathol125:221, 1986
4. Hsu SM, Hsu PL, Mulshine J, Ho YS, Ge ZH, Huang L:
Ultrastructural and biochemical studies of Leu Ml antigens ingranulocytes and H-RS cells in Hodgkin’s disease. J Natl CancerInst 77:363, 1986
5. Hsu SM, Pescovitz MD, Hsu PL: Monoclonal antibodiesagainst SU-DHL- 1 cells stain the neoplastic cells of true histiocyticlymphoma, malignant histiocytosis and Hodgkin’s disease. Blood
68:213, 19866. Kadin ME, Asbuny AK: Long term cultures of Hodgkin’s
tissue-A morphologic and radioautographic study. Lab Invest28:181, 1973
7. Peckham Mi, Cooper EH: Proliferation characterization ofthe various classes of cells in Hodgkin’s disease. Cancer 24:135,
19698. Peckham Mi, Cooper EH: Cell proliferation in Hodgkin’s
disease. Natl Cancer Inst Monograph 36:179, 19739. Kaplan HS, Gartner S: “Sternbeng-Reed” giant cells of Hodg-
kin’s disease: Cultivation in vitro, heterotransplantation, and charac-
tenization as neoplastic macrophages. Int J Cancer 19:511, 197710. Diehl V, Kirchner HH, Burnichter H, Stein H, Fonatsch C,
Gerdes J, Schaadt M, Heit W, Unchanska-Zieglen B, Ziegler A,Heintz F, Sueno K: Characteristics of Hodgkin’s disease-derived celllines. Cancer Treat Rep 66:615, 1982
I 1 . Hsu SM, Hsu PL: Phenotypes and phorbol ester-induceddifferentiation of human histiocytic lymphoma cell lines (U-937 and
SU-DHL-l) and Reed-Sternberg cells. Am J Pathol 122:223, 198612. Drexler HG, Gaedicke G, Lok MS. Diehl V, Minowada J:
Hodgkin’s disease derived cell lines HDLM-2 and L-428: Compani-son of morphology, immunological and isoenzyme profiles. Leuk Res10:487, 1986
13. Knupen K, Goldstein R, Hsu PL, Crawford Y, Zhao X, DuncM, Hsu SM: Aberrant T cell marker expression and T cell generearrangements in lymphoma/leukemia cell lines derived from
myelocyte/monocyte/histiocyte. Clin Res 35:459A, 198714. Krupen K, Zhao X, Brown L, Lachman L, Hsu SM: Expres-
sion of interleukin-I in Reed-Sternberg cells. Clin Res 35:609A,1987
15. Hsu SM, Zhao X, Hsu PL, Lok MS: Extracellulan matrixdoes not induce the proliferation, but promotes the differentiation of
Hodgkin’s cell line HDLM-l. Am J Pathol 127:9, 198716. Kamesaki H, Fukuhara 5, Tatsumi E, Uchino H, Yamabe H,
Miwa H, Shigeru 5, Hatanaka M, Honjo T: Cytochemical, immuno-
logic, chromosomal, and molecular genetic analysis of a novel cellline derived from Hodgkin’s disease. Blood 68:285, 1986
17. Hecht iT, Longo DL, Cossman J, Bolen JB, Hsu SM, IsraelM, Fisher RI: Production and characterization of a monoclonalantibody that binds Reed-Sternberg cells. J Immunol 134:4231,
198518. Hsu SM, Ho YS, Hsu PL: Expression of activated intendigi-
tating reticulum cell antigen (IRac) by H-RS cells. Lab Invest
54:27, 198619. Hsu SM, Soban E: Color modification of diaminobenzidine
(DAB) precipitation by metallic ions and its application for doubleimmunohistochemistny. J Histochem Cytochem 30:1079, 1982
20. Hsu SM, Raine L, Fanger H: Use of avidin-biotin-peroxidase
complex (ABC) in immunoperoxidase techniques: A comparisonbetween ABC and unlabeled antibody (PAP) procedures. J Histo-
chem Cytochem 29:577, 198121. Gerdes J, van Barrlen J, Pileni 5, Schwarting R, van Unnik
JAM, Stein H: Tumor cell growth fraction in Hodgkin’s disease. Am
J Pathol 128:390, 1987
22. Ho YS, Chaknaborty 5, Hsu SM: Induction of differentiation
of Burkitt’s lymphoma cell lines by phonbol ester: Possible relation-ship with pre-B cells or immature B cells. Cancer Invest 5:101,I987
23. Benson MC, McDougal DC, Coffy DS: The application ofperpendicular and forward light scatter to assess nuclear and cellular
morphology. Cytometry 5:5 15, 1984
24. Kinsella Ti, Mitchell JB, Russo A, Aiken M, Morstyn G,Hsu SM, Rowland i, Goldstein E: Continuous intravenous infusionsof bromodeoxyunidine as a clinical radiosensitizer. i Clin Oncol2:1144, 1984
25. Morstyn G, Kinsella T, Hsu SM, Russo A, Gratzner H,Mitchell 3: Identification of bromodeoxyunidine in malignant and
normal cells following therapy: Relationship to complications. mt i
RadOncol 10:1441, 1984
26. Morstyn G, Hsu SM, Kinsella T, Gratzner H, Russo A,Mitchell JB: Bnomodeoxyunidine in tumouns and chromosomesdetected with a monoclonal antibody. i Clin Invest 72:1844, 1983
27. Tomei ID, Cheney JC, Wenner CE: The effects of phorbolesters on the proliferation of C3H-1OTI/2 mouse fibroblasts: Con-sideration of both stimulatory and inhibitory effects. J Cell Physiol107:385, 1981
28. Kinzel V, Richards J, Stohr M: Early effects of the tumor-promoting phorbol ester 12-O-tetnadecanoylphonbol- 13 acetate on
the cell cycle transverse of asynchronous HeLa cells. Cancer Res41:300, 1981
29. Schwab U, Stein H, Gendes J, Lemke H, Kirchner H,
Schaadt H, Diehl VL: Production of a monoclonal antibody specificfor Hodgkin’s and Sternberg-Reed cells of Hodgkin’s disease and asubset of normal lymphoic cells. Nature 299:65, 1982
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom
1988 71: 1382-1389
SM Hsu, X Zhao, S Chakraborty, YF Liu, J Whang-Peng, MS Lok and S Fukuhara cells in culturenot actively replicating: lack of bromodeoxyuridine uptake by multinuclear Reed-Sternberg cells in Hodgkin's cell lines HDLM, L-428, and KM-H2 are
http://www.bloodjournal.org/content/71/5/1382.full.htmlUpdated information and services can be found at:
Articles on similar topics can be found in the following Blood collections
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:
Copyright 2011 by The American Society of Hematology; all rights reserved.Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of
For personal use only.on April 5, 2019. by guest www.bloodjournal.orgFrom