Cells Chronic Lymphocytic Leukemia and ...

Induction of Features Characteristic of Hairy Cell Leukemia in Chronic Lymphocytic Leukemia and Prolymphocytic Leukemia Cells H. W. Löms Ziegler-Heitbrock, Reinhold Munker, Bernd Dörken, et al. Cancer Res 1986;46:2172-2178. Published online April 1, 1986.

Updated Version

E-mail alerts Reprints and Subscriptions Permissions

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/46/4_Part_2/2172

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 July 10, 2011 Copyright © 1986 American Association for Cancer Research

(CANCER RESEARCH 46, 21 72-21 78, April 19861

Induction of Features Characteristic of Hairy Cell Leukemia in Chronic Lymphocytic Leukemia and Prolymphocytic Leukemia Cells' H.W.LÃ¶ms Ziegler-Heitbrock,2 Reinhold Munker,BerndDÃ¶rken, GerhardGaedicke, andEckhardThiel Institutefor Immunology, University ofMunich, Goethestrasse 31, 8 Munich 2 [H. W. L. Z-H.J; lnstitutefor Hematology, GSF, Landwehrstrasse 61, 8 Munich 2 (R. M., E.

T.J;

Medizinische

Polikiinik,

University

ofHeidelberg,

Hospitalstrasse

3,

69

Heidelberg[B.

D.J;

andZentrumfurKinderheilkunde,

University

ofUlm,

Prittwitzstrasse

43. 79 Ulm (G. G.J, Federal Republic of Germany

ABSTRACT

differentiation to convert cells of, e.g., CLL type of leukemic B-cell lymphoma into HCL-type cells. Previous studies dem In vitro maturation was induced with 12-O-tetradecanoylphorbol-13- onstrated that in CLL clg and immunoglobulin secretion can acetate in leukemiccell samples from patients with chronic lymphocytic be induced, which suggests maturation towards plasmacytoma leukemia(n 10) and prolymphocyticleukemia (n = 4). The cells were type cells (7). Although these findings are not disputed, we and studied for morphology, for immunologicalmarkers using the fluores cence activated cell sorter, and for acid phosphatase isoenzymesusing others (8, 9) could previously demonstrate that, by detailed analysis, patterns emerged which are consistent with differen both cytochemistryand isoelectrofocusing. Morphologically the induced changes included appearance of cells tiation of CLL cells to HCL-type cells rather than to plasma with an excentric nucleus and basophilic cytoplasm and eventually of cytoma cells. In the present report we have extended this cells with many fine cytoplasmic projections (â€œhairsâ€•). Analysis of im analysis by using flow cytometry and isoelectrofocusing on both munological

markers

by flow cytometry

revealed

that the monoclonal

CLLandPLLleukemic lymphoma cells.Ourfindings lend

antibody defined cell surface molecule HD6 (CD22), which is strongly expressed on hairy cell leukemia (HCL) but absent from plasmacytoma and plasma cells, can be induced or enhanced in the leukemic samples. In the study of acid phosphatase isoenzymes using cytochemistry we observedthe inductionof the tartrate resistant isoenzyme.Further, using

further support to the notion that CLL and PLL cells can be induced to develop into cells which are closely related to HCL cells.

isoelectrofocusing

MATERIALS

we could demonstrate

the induction of the same band

of tartrate resistant acid phosphatase with an isoelectric point of 9.0â€”9.7 as detected

also in HCL. This particular

isoenzyme

is considered

char

acteristic of HCL but is absent in plasmacytoma.Our data demonstrate that chronic lymphocytic leukemia and prolymphocyticleukemia cells can be induced to realize a common genetic program which bears char acteristics of HCL, indicating closely related than previously

that these three entities thought.

are much more

ing; B. Emmerich,

Krankenhaus

rechts

der Isar, and W. Siegert,

Kli

had not received specific cytostatic therapy for at least 4 weeks prior to

blood sampling. Patients with B-CLL are characterized by blood and bone marrow lymphocytosis. The B-CLL cells showed dense chromatin, little cytoplasm, and faint slg staining. In B-PLL, splenomegaly in the absence

Leukemia cells can be addressed as clonal transformants of the lymphohematopoietic system. Heterogeneity of leukemias reflects the different lineages, e.g., T, B, myeloid, monocytic, and the different stages of differentiation within these lineages. Classification of human B-cell tumors has made considerable progress in recent years, due to the detailed analysis of immu nological characteristics of the tumor cells (1â€”4).New stimulus to this analysis came from (a) the generation of a variety of monoclonal antibodies against cell surface molecules of human B-cells and (b) the use ofstudies involving in vitro differentiation induction of normal and malignant cells. Within the B-cell lineage HCL3 appears to be an unique type, which does not fit unconstrained into current differentiation schemes (5) and the normal counterpart of which remains to be clearly identified, although recent evidence suggests that normal sIgG' cells from peripheral blood resemble HCL cells (6). If the HCL cell is directly related to other clonal transformants of the B-cell lineage then it should be possible by in vitro induction of Received 8/6/85; revised I 2/30/85: accepted I 2/31/85. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. by grants

from

the

VW-Stiftung

Germany) and the Wilhelm-Sander-Stiftung many.

(Hannover,

Federal

Republic

of

(Munich, Federal Republic of Ger

used are: HCL,

hairycell

leukemia;

of enlarged

ACP,

lymph

nodes was the prominent

feature.

B-PLL

cells are large (>1 2 jzm) with a single prominent nucleolus and with strong slg expression

(10).

Cell Isolation. Peripheral blood mononuclear cells from these pa tients and from healthy controls were isolated from heparinized blood specimens by density gradient separation and stored in liquid nitrogen after freezing in the presence of 10% dimethyl sulfoxide using a pro grammable PTC 200 machine (Planer, Ltd., Sunberry-on-Thames,

United Kingdom). Directly before use cells were rapidly thawed in a 3TC water bath, washed twice, and resuspended in complete medium (RPMI 1640 with penicillin, streptomycin, L-glutamine, and 10% heat inactivated fetal calf serum). In Vitro Culture. Ten-mi volumes with cells adjusted to 2 X I0@/ml were cultured at 37C in a humidified 7% CO2 incubator in upright No. 3013 tissue culture flasks (Falcon, Oxnard, CA) for up to 5 days. TPA (Sigma, Munich, Federal Republic of Germany) was dissolved in ethyl acetate (Merck AG, Darmstadt, Federal Republic of Germany) at 4 mg/mI and stored at â€”20'C. Directly before use a 25-al aliquot was taken, the solvent was allowed to evaporate, and the TPA was

redissolved in 100 @zI of acetone to be diluted with 10 ml of complete medium. This stock was used for I week. After addition to cultures the final concentration

was 160 nM TPA and 0.01 % acetone.

Controls

were

set up with the same amount of acetone. Cells were recovered on days

3 and 5 and the cell number and the percentage of enlarged cells were determined

in a hematocytometer,

enlarged

cells being defined as >10

@m for CLL samples and >15 @iml for PLL samples. The cells were washed once and were used together with uncultured cells for all subsequent

2To whom requests for reprints should be addressed. 3 The abbreviations

Patients and Diagnosis.Peripheral blood samples from patients were obtained through the courtesy ofDrs. H. Theml, Krankenhaus Schwab nikum Grosshadern, Munich, Federal Republic of Germany. Patients

INTRODUCTION

I Supported

AND METhODS

analyses.

Data

from days 3 and 5 were comparable

and

were combined for presentation.

acid phosphatase;

CLL, chronic lymphocytic leukemia; FACS, fluorescence activated cell sorter; IEF, isoelectrofocusing: IP, isoelectric point: PLL. prolymphocytic leukemia: TPA, I 2-O-tetradecanoylphorbol-13-acetate: slg, surface immunoglobulin; clg, cytoplasmic immunoglobulin; MAB, monoclonal antibody; FITC, fluorescein isothiocyanate.

l3HlThymidine Incorporation. Cultured cells were adjusted to 5 x 105/ml and 100 z1 each were seeded in triplicate wells of round

bottomed microtiter plates (No. 163320; Nunc, Roskilde, Denmark). Cells were pulsed for 6 h with [3Hjthymidine (1 @tCi/well;approximately

2172


HCL PHENOTYPE 80 Ci/mmol;

NET

027Z;

New England

Nuclear,

INDUCTION

Dreieich,

IN CLL AND PLL CELLS

detect a very faint staining of CLL cell cytoplasm, which was, however, scored negative in comparison to the bright staining seen, for instance, in plasma cells. The immunoglobulin isotype of sig was determined by

Federal

Republic of Germany). They were harvested on glass fiber filters using an automated Titertek cell harvester and filters were counted in a beta counter after immersion in scintillation fluid. Cytochemistry. Acid phosphatase staining was performed

on cytospin

preparations fixed with formalin according to the method of Barka and Anderson (1 1). Tartrate resistance of the enzyme was tested in the presence

of tartrate

(2 mg/ml;

No. 804; Merck)

after readjusting

direct immunofluorescence using FITC coupled F(Ab)2 fragments spe cific for human IgG, IgM, K and X, respectively (Dakopatts, Hamburg, Federal Republic of Germany). For statistical analysis Student's t test was used.

the

pH to 5. Isoelectrofocusing.

Samples

to be analyzed

RESULTS

in IEF were washed twice

in phosphate buffered saline and the cell pellets were stored at â€”70'C. Thawed material was resuspended and adjusted to 108 cells/ml and processed as described in detail elsewhere (12). Briefly, cells were homogenized by sonic disruption (30 s, 20 W). The 20,000 x g supernatant was made I % with Triton X-100 (30 mm at room temper ature). This crude extract was stored at â€”70'C without loss of ACP activity. Isoenzyme analysis was performed by isoelectrofocusing on

horizontal thin layer slabs as described. For the visualization of the isoenzymes the gels were immersed for 3 h at room temperature into a substrate solution corresponding to that used for the cytochemical reaction. For enzyme inhibition 75 mM tartrate was added to this

staining solution. The isoelectric point of tartrate resistant isoenzyme was determined

from the pH gradient

of the IEF gels.

Immunofluorescence.All immunofluorescence procedures were per formed in immunofluorescence buffer consisting of phosphate buffered saline, 2.5% fetal calf serum, and 0.02 NaN3. Stained cells were ana lyzed after fixation with 1% paraformaldehyde (13) by flow cytometry

using a FACS-Il (Becton-Dickinson, Mountain View, CA). The FACS was operated

at 488 nm excitation

wavelength

with the photomultiplier

tube at 650â€”800V. Fluorescence intensity was expressed on a scale of

256 channels with the median representing the channel that covers 50% of the cells. These calculations

were performed

100 module. Indirect immunofluorescence

with a nuclear data ND

was used for the mouse monoclonal

antibodies directed against lymphocytecell surface determinants (Table 2). These

MABs were OKT3

(14), directed

against

T-cells,

and 63D3

(15), which recognizes monocytes and is used as an isotype control (IgGI) for HD6. HD6 reacts most strongly with HCL cells and less intensely with certain B-cell lymphomas and with a B-cell subpopula tion (16). The HD6 MAB was recently assigned to the CD22 cluster of leukocyte differentiation antigens (17). The reagents were used at

optimal concentrations for 30 mm on ice, followed by two washes and incubation

on ice with goat anti-mouse

immunoglobulin

FITC (Tago,

Burlingame, CA) diluted 1:50. Direct immunofluorescence was used for determination of slg and clg. Suspension cells or acetone-fixed cytospin preparations were in cubated for 30 mm at 4'C with goat anti-human immunoglobulin F(Ab)2 FITC (Kallestad, Austin, TX) diluted 1:32, followed by three washes before microscopic analysis. Using this procedure we could

Table 1 Basicdata on the leukemic lymphomas

studiedType

x 10-'@

of leukemiaPatientSexAge (%)CLLHa

9PLLWen

Characteristics of the Leukemic Cells. Listed in Table 1 are patients grouped according to diagnosis. All patients were over the age of 50 years and in more advanced stages of disease. Leukocyte counts were beyond 50,000/mm3 in all CLL patients and in 3 of 4 of the PLL patients. Concomitant with the high leukemic cell counts, the nonleukemic cells represented by the OKT3-positive T-cells usually were below 10% with an average percentage of 5. 1%. In all samples that could be evaluated the leukemic B-cells were monoclonal with respect to light chain expression, while for 5 cases of CLL the slg density was too low to allow for such analysis (which was performed with conventional fluorescence microscopy in this study). Staining of the clg showed no or only faint fluorescence of cytoplasm. In cytomorphology diagnosis was mostly unequivocal. In 9 of 10 CLLs the cells were small with dense chromatin and little cytoplasm, and in PLL the large cells exhibited a prominent single nucleolus. Only one case of CLL exhibited 10% cells with lymphoplasmacytoid features, i.e., an excentric nucleus and basophilic cytoplasm (patient Sc). slg on more than 60% of the cells in a sample was detected in 7 of 10 CLLs and in all PLLs. In CLL slg density, when expressed in arbitrary median channels according to FACS, gave an average of 14 median channels; only one case (Ha) exhibited a median of 72 channels. By contrast, PLL cells exhibited high slg densities with an average of 160 median channels. For HD6, conventional fluorescence microscopy de tected 12 and 21 % positive cells in 2 samples of CLL (patients Sc and Hu) but no staining in the remaining 8 samples. By contrast, FACS analysis revealed low percentages of faintly positive cells in 6 additional cases of CLL (Table 2, Column 2). For the PLL cells fluorescence microscopy and FACS gave similar results and more than 60% of the cells were positive for HD6 in 3 of 3 PLL samples tested. Intensity of staining was very weak in CLL (median, 0.6) and stronger in PLL (median, 5.3)(Table 2, Column 6). The average median for HD6 staining in 4 cases ofHCL was 11.0 Â±6.5 (SD).

(yr)StageaLeukocytes

mm3slg

isotypeslg

density'@ (OKT3 (median channels)Non-leukemic positive) T-cells

Be Hu

M M

78 60

0 II

171 84

â€”,â€”

Ku

M

78

IV

260

u,K

16

Sc

M

70

II

83

HU

M

72

II

72

â€”,â€” â€”,â€”

Lo Ne We HeM

M M M M63

69 63 79 630

II II II IV130

69 99 71 70u,K

4 2 11 I 19 103â€•

Th Huh

M F

77 75 65IV

II II III288

222 14 57u,y,K

BerM F59 a Stages accordingto the system of Rai et al. (33).

b Average leukocyte count for all patients was 120.7 Â±82.8 (SD) X I0-'/mm3. CAverage medians were 14 Â±21 for all CLL patients and 160 Â± 105 for all PLL d Average percentage of T-cells for all patients was 5.1 Â±2.9.

y,K

â€”,â€” â€”,â€”

u,y,K â€”,K72 u,y,K y,K u>250

3 3

>250 89 503

patients.

2173 Downloaded from cancerres.aacrjournals.org on July 10, 2011 Copyright © 1986 American Association for Cancer Research

2 5

2 10 4 3 4 6

5 11 4

HCL PHENOTYPE

INDUCTION


lymphomas% Table 2 TPA induced changes in HD6 binding ofleukemic B-cell (median)ControllO%Uncul stainedIntensity

of staming ControlTPA

channelsDonorUnculturedculture cultureinducedturedculturetureinducedCLLHa8@12

cul

3

TPA

50004Be229 42115Hu3011 75216Ku268

39113Sc 011

25/10Lo01 HU35 34003Ne58 27013We156

27 12â€˜1

2

02

03

8010He157 31114Mean l.6aPLLWenND40 Â±SD15.6

Â±3.5

Â±11.87.5

Â±0.70.8

34.5 Â±18â€¢1b0.6

Â±0.63.3

Â±

49ND66Th 41Ber659 Huh77 6451 69325Mean 2.7HCL(n=4)MeanÂ±SD48.5Â± Â±5.930.0 Â±SD68.7

76 310/3

20 Â±16.5

â€˜8

54

311 Â±2.13.8

56.3 Â±17.65.3

Â±1.56.5

Â±

19.411.0Â±6.5

a Direct inspectionusing the fluorescencemicroscoperevealedpositivity in only 2 cases(patients Hu and Sc). b Significant

compared

to fresh cells with P < 0.01.

C Significant

compared

to

fresh

cells

with

P

2) were those with cells@3HJThy changes in leukemic lymphoma Table 3TPA induced ofType

Appearance

midine incorpo rationLargeAcidleukemiatient(cpm)cells ofPa phosphataseCLLHa Be

@

+

75d

Hu

342

60

Ku Sc

32 2,119

29 69

HU Lo

ND 92

53 80

Ne

104 657 4,61520

We +PLLWen

(%)clg

HeNDC

higher T-cell contamination of the leukemia cell population (cf Table 3, Footnote d). Morphologically the CLL cells changed to cells with an excentric nucleus and basophilic cytoplasm with interspersed plasma cell-type cells and eventually to cells with several short (spike-like or hairy) projections. In PLL changes included de velopment of vacuoles, increase in basophilicity, and develop ment of many cellular projections. In one case each of CLL and PLL, we were able to detect cells with such â€œhairsâ€• with a frequency higher than 25% (patients Hu and Th; cf@Fig. 1), while in other samples less than 3% were identified as hairy cell-type cells according to this criterion in Pappenheim stains. clg could be induced in most samples of CLL (Table 3) while slg decreased in both control and TPA cultures to the same extent, as evidenced by FACS (not shown). The changes in HD6 staining induced by TPA in the CLL and PLL cells are summarized in Table 2. As already discussed, uncultured CLL cells showed low levels or no staining for HD6, the intensity of which usually did not exceed 1 median channel according to FACS. Control culture without TPA resulted in a decrease in the percentage of cells stained. After culture in the

+ +

+ +

â€”

+

+ +

+ +

72 17 61@a

ND

+

â€”

+

63 55 35â€”

ND

+

â€”

+

â€”

+

â€”

â€”

Th Huh Ber136

@

a

8.014 14,589 492

more than 10% of cells with bright cytoplasmic immunofluorescence

induced.

@

a

morethan 10%of cellswith intensestaininginduced;the percentage of

acid phosphatase positive cells after TPA treatment was greater than 50% in 12 samples. C ND,

d Given

not

di..

determined.

are the cpm minus the control

Â± I 210 cpm for all samples.

TPA

values, which were on the average 826

induced

[3Hjthymidine

incorporation

in

mononuclear cells from 2 healthy donors was 9250 Â±8380 cpm. The responsive cases (Sc, He, Th, Huh) were those with significantly higher T-cell contamination (8.8% OKT3 positive cells) compared to the unresponsive cases (3.6% OKT3 positive cells, P < 0.05) (cf Table I). 2174

L. Fig. 1. Morphologyof a TPA treated CLL cell.


HCL PHENOTYPE

INDUCTION

IN CLL AND PLL CELLS Table 4 TPA induction ofa tartrate resistant phosphatase as detected by cytochemistry

presence of TPA there was a striking increase of HD6 staining as evidenced by FACS analysis (Fig. 2). TPA induced or en hanced HD6 staining of the CLL cells both in percentage of positive cells (average, 34.5%) and in intensity (average median, 3.3) (Table 2). An increase by more than 10% of cells stained for HD6 was seen in 8 of 10 CLLs and by more than 20% in 5 of 10 CLLs. Intensity increased by 3 channels or more in 5 of 10 cases. In PLL, where HD6 expression was stronger to begin with, staining decreased in most control culture samples (only acetone added). TPA induced an increase by 3 channels in one case and in another by 2 channels. Finally when analyzing the effect of TPA on PLL cells in comparison to control cultures there was an increase in HD6 staining induced by TPA in 3 of

positivefor tartrate resist (%)CLL Type 2)Uncultured ofcell cultureCells

20.5a Mean

Controlcultured TPA cultured0

ant ACP

(n = 5) 1.4+ l.5a 70.8 + 9.6

PLL (n = 0 1.5 +0.7 67.5 +

Â±SD.

4 samples.

@

Thus TPA is able to induce or enhance HD6 staining in terms of both fluorescence intensity and percentage of positive cells in most samples of CLL. PLL cells, with higher staining to begin with, show a culture induced decrease which is over come by the addition of TPA in some cases. Isotype control for IgGl demonstrated no induction of binding of 63D3 to TPA treated leukemia cells in that control cultured samples gave 0.5% positives with a median of â€”0.6and TPA cultures gave 0.2% positives with a median of 0.2 (n = 3). Acid Phosphatase Analysis after Culture with TPA. Since the morphology after TPA treatment and the induction of HD6 are suggestive of a phenotype of HCL, we studied an isoenzyme characteristic of HCL, namely the tartrate resistant acid phos phatase. Acid phosphatase could be induced after TPA culture (Table 3) in all samples of CLL and in 3 of 4 PLL. In 12 of these 13 cases the percentage of acid phosphatase positive cells after TPA exceeded 50%. Tartrate resistance was tested in cytochemistry in 5 samples of CLL and in 2 samples of PLL. In every instance the staining was found to be tartrate resistant (cf Table 4 and Fig. 3), while staining

,@ -@

:.4.

-â€”@

3 Fig. 3. Cytochemical staining for tartrate resistant ACP in a TPA cultured CLL. Strong perinuclear staining (arrow) is seen in a cell with â€œhairyâ€• morphol ogy.

of normal mononuclear

TR

cells was inhibited by tartrate. The faint, scanty staining seen in uncultured

samples of leukemic lymphoma

AcP

was also sensitive

to tartrate. Since there is more than one tartrate resistant isoenzyme of ACP (18), we used IEF in order to investigate whether the enzyme induced by TPA in CLL and PLL and the ACP detected in HCL are identical with respect to isoelectric point. An

@

â€¢1

G) @0

@

E z

â€˜ â€”@wâ€•@?g@

@a

.@\

I

b4

0 Fig. 4. Isoelectrofocusing of ACP from control and TPA cultured cells. The TPA cultured cells (Lane a) exhibit an additional ACP band in the basic (upper) portion of the gel, which is absent in control cultured cells (Lane b). TR ACP, tartrate resistant ACP; I, ACP type I.

FLuorescenceIntensity Fig. 2. Effect ofTPA on cell surface staining with HD6 monoclonal antibody. Given is fluorescence intensity (linear scale) versus cell number. Superimposed curves on the left, staining of uncultured and control cultured cells; right curve, staining of the TPA treated cells (patient Be).

example of the IEF pattern induced by TPA in cells from a case with PLL is given in Fig. 4. While ACP activity is detected

2175


HCL PHENOTYPE

INDUCTION


DISCUSSION

with a major band in the acidic lower portion of the gel in both

TPAandcontrolculturesamples, additional stainingisinduced

In the present report we have analyzed two types of leukemic B-cell lymphoma, i.e., CLL and PLL. The classification was based on clinical findings, on cytomorphology, on distribution and density of immunoglobulin, and on restriction to K or A chains. Isotype analysis ofslg demonstrated monoclonality with respect to light chains in 9 cases. In S samples of CLL this determination was not possible, which is reflective of the low slg density in CLL compared to PLL as quantitated by FACS analysis. In order to avoid significant T-cell contamination in PLL.Intwoinstances thecontrolculturedcellsalsoexhibited our samples, we preferentially used cases with high leukocyte a low degree ofthis activity (Fig. 5, Lanes 2, 5). The same bands counts, such that the average T-cell percentage was 5.1%. Still, at an IP of 9.0â€”9.7as in TPA induced CLL and PLL cells are analysis of [3Hjthymidine incorporation data revealed that the also found in uncultured samples from HCL (Fig. 5A, Lane 9). autologous T-cells might be somehow involved in proliferation No cytochemical staining for ACP and no band with a basic IP seen in some instances. Saiki et a!. (19) reported on the allo in IEF were detected with TPA culture cells from a centrocytic geneic T-cell dependence of CLL B-cell proliferation. Whether lymphoma and from purified T-cells from two donors (data not in our samples T-cells give help to the monoclonal B-cells to shown). We then tested the tartrate resistance of the ACP proliferate or proliferate themselves remains to be shown. Fur directly on the gel. Incubation of the gel in the presence of ther, it is not excluded that T-cells help in the TPA induced tartrate during the enzyme staining procedure prevented almost differentiation of leukemic B-cells. all staining of ACP in the range of IP 4.5â€”6.8(Fig. SB). In vitro differentiation in B-cell tumors can be induced by a By contrast, the bands at IP 9.0â€”9.7found in TPA treated variety of substances, e.g., lipopolysaccharide, poke-weed mi CLL and PLL cells and in uncultured HCL were virtually togen, and TPA (7, 20â€”23).In previous studies using TPA, it unaffected. These experiments directly demonstrate that the IP could be demonstrated that CLL cells can be driven to accu 9.0â€”9.7bands represent the tartrate resistant ACP isoenzyme mulate clg (heavy and light chain), to synthesize J-chain (24), to increase mRNA for secretory 1gM over mRNA for membrane characteristic of HCL. by TPA in the basic portion (Fig. 4, Lane a). Further results are shown as schematic drawings (Fig. 5). In Fig. 5A two cases of CLL and one PLL are given. Again ACP activity with an IP of 4.8 to 6.8 is detected in control and TPA cultures, although control cultured cells analyzed in cytochemistry never gave a significant ACP staining, in that only few cells (less than 5%) contained a few faint granules. ACP activity with an IP of 9.0â€” 9.7 was found in TPA treated cells of the two CLLs and of one

A

pH

In

.9

AcP

-8

.7 @

.---

I â€”

â€”

.....

â€”

â€”

â€”

I = z =

@6 .5

.4

E:J @

@

Fig. 5. A, schematic drawing of isoelectro focusing analysis of ACP from 3 leukemic samples. After culture with TPA an ACP with an IP of around pH 9 is induced; eventually a faint staining is also seen in control cultures.

D@

2

EEJ

PIL tâ€”â€”â€”â€”â€”â€”â€” IPAPU. ACPLL 0Qj

E3@ I

EiJ Qj

EiJ

@J EEJ

@0â€•a@â€•

: H@.

:

AC

E@

T@@Lt@

?C@L

%

AC, acetone;for control the solvent for TPA was added to cultures. TR, tartrate resistant.

B, direct demonstration of tartrate resistance of ACP on the gel. From a duplicate run of IEF one gel was developed without (left) and one with tartrate (right) in the reaction mix ture. Only the band in the range of pH 9 is preserved in the presence of tartrate.

pH

B

â€”

TR@

AcP

â€” â€”

9 8 7

I =

I =

S 5

T = 4 @

D:J Hc@

0

rn cij:

m @L

T@

rn

rn

@LL T@

2176


Origin

HCL PHENOTYPE

INDUCTION

1gM (25), and to increase the amount of mitochondria and endoplasmic reticulum. Further, morphologically plasmacytoid features and rare plasma cells were observed (7), suggesting maturation towards plasma cells. Our results, by contrast, dem onstrate that CLL and PLL cells can be induced to acquire hairy projections, suggesting maturation to HCL-type cells. The fact that, in spite of extensive studies with TPA in CLL, the hairy cell phenotype was not observed might be explained by the low frequency of leukemic samples that show such a readily detectable .morphological change in stained smears. Support to our concept comes from the appearance of the HD6 defined antigen after TPA induction of differentiation. The HD6 MAB CD22 was produced against HCL cells, defines a cell surface molecule with an apparent molecular weight of 130,000 and 140,000 (17), and most intensively stains HCL but also, although less intensely, PLL cells, centrocyte derived B-cell tumors, and a subpopulation of normal B-cells. No cell surface staining of plasmacytomas is observed (16). In CLL only a few cases stain with this reagent in a low percentage of cells. In the present report we could detect by FACS some additional cases with faint staining not detectable by conven tional fluorescence microscopy. After culture in the presence of TPA the cell surface staining for HD6 was either induced or enhanced in CLL and PLL in terms of both percentage of positive cells and intensity of staining. It could be argued that the increase in intensity of staining as detected by FACS merely reflects the increase in cell surface area with a constant density of binding sites. Although determinations of cell surface area have not been performed, our fluorescence microscopy read ings, which were always done in parallel to FACS analysis, also revealed an overt increase in intensity of staining, thus sup porting the assumption that TPA induces not only a higher number but also a higher density of HD6 defined cell surface molecules in CLL and PLL. Two other MABs with a specificity similar to that of HD6, namely RFA-4 (26) and FMC7 (27), have been reported. Mo lecular weights of the antigens recognized are not published, such that a comparison with HD6 remains incomplete. Induc tion of FMC7 has been reported in the study of a single case of CLL (23). This finding is in concert with our results; it was, however, interpreted as maturation towards immunocytoma- or PLL-type cells. Still, the appearance of a cell surface marker which is absent from plasmacytoma cells argues against differ entiation to the level of the plasma cell. Accumulation of clg is not in conflict with the concept of maturation towards HCL type cells, since some cases of HCL have been demonstrated to contain clg (5). Further support to our conept of induction of a phenotype similar to HCL comes from the isoenzyme studies. The cyto chemical staining revealed the appearance, in the majority of cells in almost every sample tested, of a tartrate resistant acid phosphatase, as reported earlier (8, 9). Using IEF we were able to demonstrate the induction by TPA of an isoenzyme similar to HCL in the leukemic cells from CLL and PLL patients. This band tended to be broad with an IP of 9.0â€”9.7.Two to 3 faint lines within this band indicate that there might be microheter ogeneity also for this type of isoenzyme, as is seen for instance for isoenzyme


tartrate resistant. The detection also in control cultured CLL cells of some tartrate resistant ACP activity by IEF but not by routine cyto chemistry underlines the power of the IEF approach for detec tion of this enzyme. The appearance of tartrate resistant ACP, however, is not a general phenomenon observed after culture

(Â±TPA) ofleukemicornonleukemic lymphocytes, sincenormal T-cells and a case of centrocytic-centroblastic lymphoma did not exhibit this isoenzyme after such treatment. We also oh served isoenzymes with IP of 4.5â€”6.8in both TPA and control cultures and eventually uncultured cells. This activity in our hands was not detected in cytochemical stains of control cul tures and uncultured cells (data not shown), which is probably also due to different sensitivities of the methods as discussed (12). Ofthe 14 cases in the present report only one was unrespon sive to TPA with respect to every parameter tested. Although this leukemia sample was a PLL, our data, which demonstrate responsiveness in the other 3 PLL, do not support the conclu sion by Okamura et a!. (21) that PLL-type cells are poor responders to TPA. Our data demonstrate that the 2 types ofleukemic lymphoma studied are capable of acquiring some features of HCL. Thus CLL and PLL are very similar with respect to the genetic program that they can realize in response to TPA and this genetic program includes features of HCL. The assignment of HCL to the B-cell lineage has been clearly established by the presence ofintrinsic slg (e.g., Ref. 28) and by the demonstration of immunoglobulin-gene rearrangement (29, 30). The precise location of this type of cell in a B-cell differentiation scheme, however, remains to be determined. Our data, showing the conversion of CLL and PLL cells into HCL-type cells, would suggest a close relationship between these 3 types of lymphoma, a conclusion also supported by the earlier finding of a tartrate resistant ACP in cases of PLL (31). Thus HCL appears to be more closely related to other malignant transformations of the B-cell lineage than previously thought. Previous studies (32) have shown that TPA can induce morphological changes in

HCLcells,whilesurfaceglycoproteinsstudiedin gelelectro phoresis did not change. It remains to be shown whether analysis with monoclonal antibodies and the use of other in ducers ofdifferentiation reveal evidence for maturation of HCL cells to plasma cell-type cells or whether HCL-type cells rep resent mature cells of a separate branch of the B-cell lineage that includes cells of the CLL type. ACKNOWLEDGMENTS We acknowledgethe excellent technical assistance ofC. Wagenpfeil, E. B. Weiss, G. Scheffel, and A. Zimmer and the provision of patient specimens

by Drs. Theml,

Emmerich,

and Siegert.

REFERENCES

I.

Development of the zymogram in the presence of tartrate revealed that the band with an IP of 9.0â€”9.7remained unaf fected, while all other isoenzymes almost completely disap peared. This demonstrated that TPA induces in CLL and PLL an ACP isoenzyme which bears characteristics ofthe isoenzyme found in HCL, in that it exhibits the same IP range and it is

1. Anderson, K. C., Bates, M. P., Slaughenhoupt, B. L., Pinkus, G. S., Schloss man, S. F., and Nadler, L. M. Expression ofhuman B cell-associated antigens on leukemias and lymphomas: a model of human B cell differentiation. Blood, 63: 1424â€”1433,1984. 2. Foon, K. A., Schroff, R. W., and Gale, R. P. Surface markers on leukemia and lymphoma cells: recent advances. Blood, 60: 1â€”19,1982. 3. Koziner, B., Kempin, S., Passe, S., Gee, T., Good, R. A., and Clarkson, B. D. Characterization of B-Cell leukemias: a tentative immunomorphological scheme. Blood, 56: 815â€”822,1980. 4. Magrath, I. T. Lymphocyte differentiation: an essential basis for the compre hension oflymphoid neoplasia. J. Natl. Cancer Inst., 67: 501â€”514,1981. 5. Jansen, J., Schuit, H. R. E., Meijer, C. J. L. M., Van Nieuwkoop, J. A., and Hijmans, W. Cell markers in hairy cell leukemia studied in cells from 51 patients. Blood, 59: 52â€”60,1982.

2177


HCL PHENOTYPE

INDUCTION


6. Machii, T., and Kitani, T. Similarities between IgG-bearing lymphocytes and

hairy cells:cytologicand cytochemicalstudies. Blood,64: 166â€”172, 1984. 7. TÃ¶tterman, T. H., Nilsson, K., and SundstrÃ¶m, C. Phorbol ester-induced differentiation of chronic lymphocytic leukaemia cells. Nature (Lond.), 288: 176â€”178, 1980. 8. Caligaris-Cappio, F., Janossy. G., Campana, D., Chilosi, M., Bergui, L., Foa. R.. Delia, D., Giubellino, M. C., Preda, P., and Gobbi, M. Lineage relation ship of chronic lymphocytic leukemia and hairy cell leukemia: studies with TPA. Leuk. Res., 8: 567â€”578,1984. 9. Ziegler-Heitbrock, H. W. L., Dorken, B., Munker, R., RiethmOller, G., Thierfelder, S., and Thiel, E. In-vitro induction ofsome features ofhairy cell leukemia in chronic lymphocytic leukemia and immunocytoma cells. Blut, 50:29â€”31. 1985. 10. Galton, D. A. G., Goldman, J. M., Wiltshaw, E., Catovsky, D., Henry, K., and Goldenberg, G. J. Prolymphocytic leukaemia. Br. J. Haematol., 27: 7â€” 23,1974. I I. Barka, 1., and Anderson, P. Y. Histochemical methods for acid phosphatases using hexazonium pararoseaniline as coupler. J. Histochem. Cytochem., 10: 741â€”753, 1962. 12. Gaedicke, G., Arndt, R., Kabisch, H., and Landbeck, G. Isoenzyme studies in human leukemia. I. Acid phosphatase. Leuk. Res., 3: 417â€”423,1979. I 3. Lanier, L. 1., and Warner, N. L. Paraformaldehyde fixation of hematopoietic cells for quantitative flow cytometry (FACS) analysis. J. lmmunol. Methods, 47:25â€”30, 1981. 14. Kung, P. C., Goldstein, G., Reinherz, E. L., and Schlossman, S. F. Monoclo nal antibodies defining distinctive human T cell surface antigens. Science (Wash. DC), 206: 347â€”349,1979. 15. Ugolini. V., Nunez, G., Smith, R. G., Stastny, P., and Capra, J. D. Initial characterization of monoclonal antibodies against human monocytes. Proc. Natl. Acad. Sci. USA, 77: 6764â€”6768,1980. 16. Dorken, B., Schwarz, E., Feller, A. C., Hammerling, G., and Hunstein, W. Production of monoclonal antibodies for the diagnosis of minimal infiltration of leukemic cells into the bone marrow: B cell specific antibodies. Verh. Dtsch. Ces. Pathol., 67: 65â€”69,1983. I 7. Moldenhauer. G., DÃ³rken, B., Schwartz, R., Pezzutto, A., and Hammerling,

G. J. Characterizationof a human B-lymphocytespecificantigen, definedby the monoclonalantibodies HD6 and HD39. ln: E. L. Reinherz, B. F. Haynes, L. M. Nadler. I. D. Bernstein (eds.), Leucocyte Typing II. Berlin: Springer Verlag, in press, 1986. 18. Yam, L. 1. Clinical significance of the human acid phosphatases. Am. J. Med., 56:604â€”616, 1974. 19. Saiki, 0., Kishimoto, T., Kuritani, T., Muraguchi, A., and Yamamura, Y. ln vitro induction of 1gM secretion and switching to lgG production in human B leukemic cells with the help of T cells. J. lmmunol., 124: 2609â€”2614, 1980. 20. Gordon, J., Mellstedt, H., Aman, P., Biberfeld, P., and Klein, G. Phenotypic

21. 22. 23.

24. 25. 26. 27. 28.

29.

modulation ofchronic lymphocytic leukemia cells by phorbol ester induction of 1gM secretion and changes in the expression of B cell-associated surface antigens. J. Immunol., 132: 541â€”547,1984. Okamura, J., Gelfand, E. W., and Letarte, M. Heterogeneity of the response of chronic lymphocytic leukemia cells to phorbol ester. Blood, 60: 1082â€” 1088,1982. Robert, K.-H., Bird, A. G., and MÃ¶ller,E. Mitogen-induced differentiation and human CLL lymphocytes to antibody-secreting cells. Scand. J. Immu nol., 10: 447â€”452,1979. Robert, K.-H., Juliussin, G., and Biberfeld, P. Chronic lymphocytic leukae mia cells activated in vitro reveal cellular changes that characterize BProlymphocytic leukaemia and immunocytoma. Scand. J. Immunol., I 7: 397â€”401,1983. Yasuda, N., Shimada, H., Kanoh, T., and Uchino, H. J chain synthesis and ultrastructural changes in chronic lymphocytic leukaemia cells during in vitro differentiation. Clin. Exp. Immunol., 57: 351â€”357,1984. Cossman, J., Neckers, L. M., Braziel, R. M., Trepel, J. B., Korsmeyer, S. J., and Bakhshi, A. In vitro enhancement of immunoglobulin gene expression in chronic lymphocytic leukemia. J. Clin. Invest., 73: 587â€”592,1984. Gobbi, M., Caligaris-Cappio. F., and Janossy, G. Normal equivalent cells of B cell malignancies: analysis with monoclonal antibodies. Br. J. Haematol., 54:393â€”403, 1983. Catovsky, D., Cherchi, M., Brooks, D., Bradley, J., and Zola, H. Heteroge neity of B-cell leukemias demonstrated by the monoclonal antibody FMC7. Blood, 58: 406â€”408,1981. Rieber, E. P., Hadam, M. R., Linke, R. P., Saal, J. G., RiethmUller, G., Von Heyden, H. W., and WaIler, H. D. Hairy cell leukaemia: surface markers and functional capacities of the leukaemic cells analysed in eight patients. Br. J. Haematol., 42: 175â€”188,1979. Cleary, M. L., Wood, G. S., Warnke, R., Chao, J., and Sklar, J. Immuno globulin gene rearrangements in hairy cell leukemia. Blood. 64: 99â€”104, 1984.

30. Korsmeyer,

S. J., Greene, W. C., Cossman,

J., Hsu, S., Jensen, J. P., Neckers,

L. M., Marshall, S. L., Bakhshi, A., Depper, J. M., Leonard, W. J., Jaffe, E. S., and Waldmann, T. A. Rearrangement and expression of immunoglobulin genes and expression of Tac antigen in hairy cell leukemia. Proc. NatI. Acad. Sci. USA, 80: 4522â€”4526,1983. 31. Katayama, I., Aiba. M., Pechet, L., Sullivan, J. L., Roberts, P., and Hum phreys, R. E. B-Lineage prolymphocytic leukemia as a distinct clinicopath ologic entity. Am. J. Pathol., 99: 399â€”412,1980. 32. Lockney, M. W., Golomb, H. M., and Dawson, G. Unique cell surface glycoprotein expression in hairy cell leukemia: effect of phorbol ester tumor promoters on surface glycoproteins, cell morphology, and adherence. Cancer Res., 42: 3724â€”3728,1982. 33. Rai, K. R., Sawitsky, A., Cronkite, E. P., Chanana, A. D., Levy, R. N., and Pasternack, B. S. Clinical staging of chronic lymphocytic leukemia. Blood, 46:219â€”234, 1975.

2178
