Interferon exerts a cytotoxic effect on primary human myeloma cells

Interferon Human

Exerts

a Cytotoxic

Myeloma STEFAN

Effect on Primary

Cells

EINHORN,

JAN-OLOF

FERNBERG,

DhN GRANDl?R

Radiumhemmet,

Karolinksa

and ROLF LEWENSOHN

Hospital,

104 01 Stockholm,

Sweden

Abstract -.4 dye exclusion method was used for testing the sensitivity of primary human ncyeloma and normal bone-marrow cells to interferon (IFh’). Following 4 days of incubation umith5000 unit.~/ml of natural (n) IFNa, myeloma cells in cultures from intermediate

there was a >90%

some patients,

or no sensitivity to IFN.

a 4-day exposure to nIFhr-a

1MO%

and non-malignant

of malignant

other patients showed

The number of viable non-malignant

(from the same patients) following Exposure

decrease in the number of viable

whereas myeloma cells from (5000

bone-marrow

bone-marrow

units/ml)

cells

decreased by

cel1.r to natural

p-

and

recombinant cy- and ?I-IFN, also induced a decrease in the number of viable cells. The decreased number of t’iable m_>‘elomacells could be observed already after 1 day of exposure to IF.+:-(Y in

vitro. To test whether inhibition of proliferation ation. as measured

by [ ‘Hjthymidine

uptake,

I ~2% of the myeloma cells were labeled

could account for the obseroed e]fects, proliferwas studied in some experiments.

\ itrv, whereas the proportion of labeled non-malignant dis was approx. 40%. induced reduction of cell number in normal bone-marrow cell multiplication

inhibitory effect of IFN,

attributed to cell multiplication

inhibition.

n!yeloma cells could be attributed

Onb approx.

with [“HI thymidine during the 4 days of culture in cells could possibly

Thus, the IRAbe attributed

to a

whereas the effect observed in myeloma cells cannot be To test

thepossibility that the reduction in the number of

to the activity of autologous

cytotoxic T-cells,

)VK-rel1.r or

macropha,ges, these cells Were depleted in some experiments. Depletion of these cells did not, howeaer, influence the IF;LT-induced decrease in the number of viable myeloma cells. I)$> thus conclude that IF.V can reduce the number of iliable tumor cells by a cytotoxic eflect, unrelated to cell multiplication inhibition.

INTRODUCTION INTERFERONS(IFNs) have the capacity the multiplication of malignant cells in z&o [ 1, 21. The direct

MATERIALS to inhibit

Patients

and non-malignant ccl1 multiplication

Bone-marrow patients

inhibitory effect of IFN has been suggested to be the mechanism by which IFN induces remissions in human

malignancies

[3 1. However,

patients

in some patients

purified [6].

another

The

derived

possibility

obtained

in Table

Data

from on

19

these

1.

(na-IFN) human

was prepared from Namalwa cells and

by an anti-IFN-antibody specific

activity

of this

affinity preparation

system was

from

E.

coli.

The

specific

activity

of this

preparation was 3.2 X lo8 units/mg of protein and the purity was >99%. E. coli derived recombinant IFN-)I (t-y-IFN; from Ernst-Boehringer-Institut fur

is su,ggcstcd, based on work using a dye exclusion assay [.jJ, namely that IFNs exert antitumor effects 1)) a direct cytotoxic effect on primary human mycl0111a

are shown

were mycloma.

2 X 10” units/mg of protein and the purity approx. 90%. Recombinant a2-IFN (rcr-IFN; from ErnstBoehringer-Institut fur Arzncmittelforschung) was

nations for this rapid effect of IFN on malignant tumors are, for instance, stimulation ofan antitumor immune response by IFN, or an induction of diffcrIn this paper

samples multiple

The natural a-IFN Sendai virus induced

occur too fast to be explained simply by an inhibition of the multiplication of tumor cells. Possible expla-

by IFN.

with

IFN preparations

\vith B-ccl1 malignancies, such as myeloma, remissions have been shown to occur within a few weeks after initiation of IFN therapy [4]; effects

entiation

AND METHODS

Arznemittelforschung) had a specific activity of 2 x lo7 units/mg of protein. Natural B-IFN (nBIFN) was produced from fibroblasts by induction with poly(1) poly(C). The specific activity of the

cells.

1505

S. Einhorn et al.

1506

Table 1. Patient data Patient

75 71 55 76 65 69 81 58 69 72 65 64 51 89 63 65 59 62 69

2 3

6 8 9 10 11 12 13 14 15 16 17 18 19

Sex

Phenotype

F F F F M F M F F ,M M M M M hl M F F M

IgGk Ig*’ Ig*k IgAl IgGk IgGk

BJl Non-S IgAl IgAl

BJk BJk IgGk

BJk BJ’

1g*k IgAk Ig*1 IgGk

partially purified preparation was 1.6 X 10” units/ mg of protein [7]. The antiviral activities of the preparations were determined bition of the cytopathogenic stomatitis

by assaying inhieffect of vesicular

virus in human fibroblasts

as previously

described [8]. The antiviral &ctivities are expressed in international units by comparison with intcrnational

reference

preparations.

bone marrow

samples

and periph-

eral blood were centrifuged on a layer of FicollHypaque [9]. The interphase cells were collected and washed twice by centrifugation in medium (RPM1 1640 with 1% L-glutamine, 10% fetal calf serum, 50 pg/ml of penicillin and 50 pg/ml of streptomycin). In some experiments monocytes and granulocytes were depleted by plastic adherence during 30 min as well as by iron powder and a magnet. High- as well as low-avidity by E-rosette sedimentation

T-cells were depleted using neuraminidase

treated sheep red blood cells as previously described [lo]. Following counting, 10” cells were added to 5ml plastic tubes together with medium with or without IFN and incubated at 37°C for l-9 days. Dye exclusion assay A dye exclusion assay originally reported by Weisenthal et al. [5] and later modified by Bird et al. [ 1 l] was used with some additional modifications. Briefly, a defined number of permanently fixed duck red blood cells (DRBC) were added to the cell suspensions after incubation, after which the cells were stained with 2% fast green and 1% nigrosin dye solutions;

Percentage plasma cells in the bone marrow

NO Chemotherapy Chemotherapy x0 Radiotherapy No Ax’o No No Chemotherapy No Chrmotherapy Chemotherapy Chemotherapy No Chemotherapy Chemotherapy Chemotherapy Chemotherapy

were

69

12 33 10 14 31 75 17 35 90 95 17 22 33 30 31 50 35 34

cytocentrifuged

0.1 ml of both. After 10 min the cells

onto

slides

at 500 rpm for

10 min after which the slides were air-dried and fixed in methanol for 20 s. All slides were count’erstained with May-Grunwald-Giemsa (MGG) to enable differentiation

between cell types.

Assay interpretation The DRBSs served this assay, minimizing

Preparation and culture of cells Heparinized

Previous treatment

as an internal standard in the problems with uneven

distribution of cells on the slides, cell autolysis and proliferation ofcells during incubation, as discussed by Weisenthal et al. [5]. Live tumor cells as well as live normal bone marrow cells were calculated in relation to DRBC in all slides. The number ofviable cells in the cultures with IFN were expressed as a percentage

of the number

of viable cells in control

cultures. Autoradiography DNA synthesis in the myeloma cells and in the normal bone-marrow cells was determined by the following procedure: a suspension of 10” cells/ml was incubated with 10 pCi [“H]thymidine/ml medium at 37°C. The incubation time was l-4 days. Following staining with MGG, autoradiography was performed with liquid emulsion technique using Ilford K2 film. The labeling index was calculated by counting 500 cells and determining the percentage of cells containing over the nucleus.

more than 10 grains

Statistical analyses Statistical significances were evaluated by Student’s t-test and by linear regression analyses.

Interferon and Myeloma Cells RESULTS In

19 myeloma

patients

In the

number

seven

different

of viable

1507 patients

with

IFN preparations

myeloma

the

effect

on the viability

of

of myel-

myeloma cells was tested following exposure to ncxIFN for 4 days. In relation to control cultures, naIFN caused a more than 90% decrease in the

oma cells and non-malignant cells was tested. In most patients, ry-IFN had the most pronounced effect on cell viability (Table 3), whereas IFN-a

number

(natural efficient

of viable

myeloma

cells

in cultures

from

three patients (Table 2). In cells from some patients, IFN-a had no major effect on the viability of the myeloma cells, whereas cells from other patients

decrease 2). As

was the least of viable cells.

The difference between the o-IFN preparations and the p- and y-IFNs in reducing the number of myeloma cells was in some casts statistically sig-

showed an intermediate sensitivity to IFN (Table 2). In contrast to the results obtained with myeloma cells, non-malignant cells from none of the patients exhibited a similar high sensitivity to nol-IFN, i.e. the greatest 63% (Table

as well as recombinant) in inhibiting the number

nificant (P < 0.01 between ro-IFN and P < 0.01 between ra-IFN and ry-IFN). The effect of various on the viability of bone

in viable cells observed was can be seen in Fig. 1, the

five patients

with

r$-IFN;

concentrations of no-IFN marrow cells was tested in

myeloma.

The

effect

of IFN

sufficient

between the IFN sensitivity of non-malignant cells and the IFN sensitivity of the myeloma cells, for

viable myeloma cells (Fig. 2). Optimal effects seen using 500-5000 IFN-(-w units/ml. In

individual patients. We studied whether

malignant

number

of mycloma

be correlated

the effects

of IFN

and non-malignant

to clinical

parameters.

on the

female patients cells from male

from this, correlations

was a

bone-marrow by

IFN-(r

culture

with

parameters such as Ig type, age, previous treatment and the proportion of plasma cells in the bone

observed already The possibility

marrow.

viable

myeloma patients.

of no-IFN

natural

patient, who showed a, a minor decrease

Table 2. Infruence of natural IFN-a

myeloma

(5000 units/ml;

and non-malignant

IFN-a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Mean + S.E.

viable cells

0 28 4 7 30 25 93 53 22 60 65 106 88 63 116 105 63 30 92 55 ? 9

the reduction

of were non-

in cell

found

to be

dose-

studies

were perfor-

(5000 units/ml)

on

(Fig. 3). In the other

after that

1 day of culture with IFN. IFN reduced the number

cells by a stimulation

4 days of culture)

on the

cells in bone-marrow from

19

The data are expressed as a percentage of control cells cultured

Percentage myeloma

number

a weaker susceptibility to IFNin myeloma cell viability was

in medium only Patient

also

in the

myeloma cell viability. In one of the patients, no major effect on cell viability was seen until day 4 of

significant and clinical

number of viable malignant

cells, was

med on the influence

were more sensitive to IFN patients (P < 0.001). Apart

there were no statistically between IFN sensitivity

a reduction

dependent (Fig. 2). In two patients, time-kinetic

statistically significant correlation between sex and the sensitivity of the myeloma cells to IFN; cells from than

to cause

number

cells could There

on

myeloma cell viability was dose-dependent (Fig. 2). In some of the patients 0.5 units/ml of no-IFN was

myeloma cells exhibited a more hcterogencous picture in regard to IFN sensitivity as compared to non-malignant cells. There was no correlation

Percentage viable non-malignant cells 48 90 67 58 65 92 42 40 37 N.1). ND. 73 N.D. a3 88 76 86 51 52 66 2 5

of cytotoxic

of

S. Einhorn et al.

1508

reduction in the number of viable non-malignant cells by 12-24% during four days of culture in vitro.

??

DISCUSSION I .

IFN

has

previously

been

shown

to inhibit

the

multiplication of normal and malignant cells in vilro cell-lines, such [ 1, 21. In some highly IFN-sensitive as L1210, the cells have been shown to die after a period of culture in vitro [ 121. In the cast of L 12 10 this effect has, however, been shown to be secondary to the slowing in the multiplication rate of the cells

Im ?? ??

[ 131. In this

work

we show

that

IFN

in vitro can

reduce the number ofviablc primary mycloma cells. This is not due to an inhibition ofcell multiplication, since only a few per cent of the cells arc labeled with [“Hlthymidine, indicating DNA synthesis, during the period of in vitro culture. The possibility that

I ”

IFN reduced the number of viable tumor cells indirectly, by an effect on immunologically active

.

0.0

Myelome cells

Non-Malignant cells

cells, is not granulocytes

FiE. 1. Percentage viable cells following incubation in the absence or presence of na-IFN (5000 units/ml) for 4 days ? ?denotes different patients.

effect.

of these

cells from

the preparations

IFN

therapy

acts

by a direct

effect

on the myeloma

was determined

in bone marrow

of the patients.

Only

l-2%

cells,

cultures

the

IFN

cytotoxic

to the results

usually

effect

obtained

caused

of IFN

been shown

only

on

in estab-

with myeloma

minor

decreases

in

number

[“Hlthymidine

of

during

4 days

in uitro.

of culture

Thus, possibly the reduction in the number ofviablr normal bone-marrow cells could be explained by a

with IFN

cell multiplication

from some

of the myeloma

A direct

of non-malignant bone-marrow cells 40% of non-malignant (Table 2). App roximately bone marrow cells were found to bc labeled with

cells

the number

myeloma cells and non-malignant cells labeled [3H] thymidine during a 4-day culture without

[3].

In contrast

did

decrease in the numwhich indicates that

(Table 4). By the use of autoradiography,

that IFN can have a direct

malignant cells has previously lished cell lines [ 141.

monocytes, granulocytes as well as high- and lowavidity T-cells from the bone-marrow preparations. not abrogate the IFN-induced ber of viable myeloma cells,

We thus conclude

cytotoxic effect on myeloma cells at concentrations that are regularly obtained in the serum during IFN

monocytes or lymphocytes, active against the autologous myeloma cells was tested by’depletion of

Depletion

likely, since depletion of monocytcs, and T-cells did not influence this IFN

inhibitory

effect of IFN. Whether

the cytotoxic effect of IFN is exclusive for myeloma cells or whether also other malignant cells arc

cells

were labeled with [“Hlthymidine during the culture period. During the same time-period ncu-IFN caused

susceptible

a reduction in myeloma in cultures from these period of time, approx.

cell number by up to 40% patients. During the same 40% of the non-malignant

investigated. OL-, /3- and y-IFN differed in their capacity to reduce the number of viable myeloma cells; p- and

cells

whereas

entered

Table 3. InJuence

S-phase,

of

various IFNs

ncu-IFN

caused

(500 units/ml)

y-IFN

a

are

to this effect

more

on the number of viable malignant

efficient

of IFN

is presently

than

and non-malignant

a-IFN

on

being

a per

cell> ,from Jeuen myrloma

patients Patient IKX-IFN

Percentage viable malignant I+-IFN I-Cl-IFN

cells ry-IFN

nCX-IFN

Pwccntagc viahlc non-malignant ra-IFN t$-1F.X

Myeloma

1

10

4 5 6 7 8 9

19 20 28 84 81 58

35 47 28 32 91 61 45

5 24 20 22 84 49 24

40 17 12 12 83 23 15

93 63 61 70 76 74 62

135 86 76 86 89 48 51

68 86 66 62 71 73 45

43 -c 12

48 2 8

33 5 IO

29 t IO

71 *4

82 5 11

67 k 5

Mean f S.E.

crlls ry-IFN

1509

Interferon and Myeloma Cells Non-Malignant Cells

Myeloma Cells

PI00

H E 8 6

80

@0

=mw

3 $

40

.!! > 20

0 5

0.5 IFN

ITi,q.

50

5000

IFN-a

concentrations

on the number

lines

denote

antiviral unit basis. As is well known, it is difficult to compare the activity of different IFNs on a per antiviral unit basis. Studies are therefore in process different

a-IFN

subtypes

myeloma

ultimately leading IFN has previously

different

oyuiable

a terminal

5000

myeloma

and

non-malrgnant

bone-mnrrou

tells

patirntr.

Table 4. Injuence

of natural IFA-a

(5000

units/ml)

on

the

number of viable myetoma cells. Tests Were done on separated cells and cells depleted of monocytes, granulocl;tes and T-cells Patient

cells,

500

differ

How does IFN cause a reduction in the number of viable myeloma cells? One possible explanation could bc that IFN induces terminal differentiation of the

50

5

IFN concentration (UlmL)

The

to evaluate whether in this function.

05

0

concentration (U/mu

ofrariousnatural

2. In/luence

500

differentiation

Percentage

viahlc

malignant

cclla

Unseparatrd

SrparattXl

13

88

4’1

14

63

03

to the death of the affected cells. been shown to induce differcntiation

in malignant

B-cells

from patients

with CLL

in vitro [ 15, 161. Another possibility is that IFN acts by influencing the expression of genes that are of importance IFN

for the viability

has been

a variety

shown

of genes

[17] and

oncogenes [18, 191. Cells from different bility

to IFN

material and

of the malignant

to influence

amongst

patients

(Table

2).

we found

a close

the sensitivity

of the

them,

rrlativcly

correlation myeloma

patients

being

cells from

patients

(P < 0.001).

not this holds being

true in a larger

investigated.

of

several

vary in their suscepti-

In our

cells from female male

cells.

the expression

cells

sex

to IFN;

more sensitive material

A variability

small

bctwccn

than

M’hether

or

is presently

in the in vitro sensi-

tivity to IFN has also been shown for primary myeloma cells using an agar-colony assay and an autoradiographic

method

measuring

inhibition

of

3H uptake [20]. In the latter study a correlation between in vitro sensitivity and the clinical response to IFN-a

1

1

2

3

4

5

6

7

8

9

Days Fig.

3.

units/ml)

Time-course

stu$

on the number

influence

of

of arable

on

the

myeloma

cells.

different

patients.

natural The

IFN--a lines

therapy

was found.

Whether

the observed

reduction in colony formation was due to the cell multiplication inhibitory effects of IFN, or whether it was due to the direct cytotoxic effect of IFN is not (5000

denote

two

known. A major drawback in optimizing IFN therapy is that we do not know how IFN exerts antitumor

S. Einhorn et al.

1510 effects.

Several

different

mechanisms

have

been

suggested to explain how IFN can induce remissions in tumors. Apart from indirect effects mediated by, for instance,

the immune

system,

stromal

this work, by another effect

on the results

we now suggest that mechanism, namely

on the tumor

cells.

Whether

by which and/or on.

IFN induces in other

remissions

malignancies,

in multiple can only be

cells or

endocrine organs, IFN may act by direct effects, such as cell multiplication inhibition or induction of differentiation..Based

means

myeloma speculated

obtained

in

IFN may also act a direct cytotoxic or not this is a

Acknowledgements-Theexcellent technicalassistanceofMiss Karin Burvall, Mrs Inger Foskolos, Mrs Lena Odin and Mr Dimitrios Papadimas is acknowledged. We thank Drs Ernest Knight Jr, Otto Adolf and David Secher for providing the IFN preparations used. This work was supported by grants from The Swedish Cancer Society, The Cancer Society of Stockholmand The King Gustav V Jubilee Fund, Stockholm.

REFERENCES 1. Paucker

2.

7. 8.

9. 10.

K, Cantell K, Henle W. Quantitative studies on viral interference in suspended L cells. III. Effect of interfering viruses and interferon on the growth rate of cells. Virolou 1962, 17, 324-334. Einhorn S, Strander H. Interferon therapy for neoplastic diseases in man. In vitro and in Workshop, W. vivo studies. In: Human Interferon. Production and Clinical Use. A Symposium Alton Jones Cell Science Center, Lake Placid, U.S.A., 19-20 May 1977, 159-174. Strander H. Interferon treatment of human neoplasia. Adv Cancer Res 1986,46, l-265. Mellstedt H, Ahre A, BjGrkholm M, Helm G, Johansson B, Strander H. Interferon therapy in myelomatosis. Lancet 1979, 1, 245-247. Weisenthal LM, Marsden JA, Dill PL, M acaluso CK. A novel dye exclusion method for testing in vitro chemosensitivity of human tumors. Cancer Res 1983, 43, 749-757. Finter NB, Fantes KH. The purity and safety of interferons prepared for clinical use. The case for lymphoblastoid interferon. In: Gresser I, ed. Interferon. New York, Academic Press, 1980, 65-80. Obert HJ. Clinical trials and pilot studies with beta-interferon in Germany. In: UCLA Symposium on Molecular and Cellular Biology. 1982, Vol. XXV, 426-432. Einhorn L, Einhorn S, Wahren B. Interferon induction in human leukocytes after in vitro exposure to cytomegalovirus or Epstein-Barr virus. Influence ofinterferon on the expression of viral antigens. Znterviroloyy 1985, 23, 10-149. Bijyum A. Separation of leukocytes from blood and bone marrow. &and J Clin Lab Invest 1968, 21, Suppl 97, I-109. Wasserman J, Einhorn S, Petrini B, von Stedingk L-V. Further studies on the influence of interferon on pokeweed mitogen induced Ig-synthesis in vitro. J Clin Lab Immunol 1985, 18,

187-189.

AG, Gilby ED. In vitro determination of tumour chemosensitivity in 11. Bird MC, Bosanquet hematological malignancies. Hematol Oncol 1985, 3, 1-9. I, Brouty-Boy6 D, Thomas M-T, Macieira-Coelho A. Interferon and cell division. 12. Gresser II. Influence of various experimental conditions on the inhibition of L1210 cell multiplication in vitro by interferon preparations. J Nat1 Cancer Inst 1970, 45, 1145-l 153. D. The use of the chemostat to study the relationship between cell 13. Tovey M, Brouty-Boy6 growth rate, viability and the effect of interferon on L1210 cells. Exp Cell Res 1979, 118, 383-388. activity of interferon-gamma and its synergism with 514. Le J, Yip YK, Vilcek J. Cytolytic fluorouracil. Int J Cancer 1984, 34, 495-500. S, Robert K-H, Ostlund L, Juliusson G, Biberfeld P. Interferon induces prolifer15. Einhorn ation and differentiation in fresh chronic lymphocytic leukemia cells. Antiviral kes 1984, 3,

40. 16. Ostlund leukemia 17.

18. 19.

20.

L, Einhorn S, Rob&t K-H, Juliusson G, Biberfeld P. Chronic B-lymphocytic cells proliferate and differentiate following exposure to interferon in vitro. Blood 1986, 67, 152-159. Friedman RL, Manly SP, McMahon M, Kerr IM, Stark GR. Transcriptional and posttranscriptional regulation of interferon-induced gene expression in human cells. Cell 1984, 38, 745-755. Jonak GJ, Knight E Jr. Selective reduction of c-myc mRNA in Daudi cells by human p interferon. Proc Nat1 Acad Sri USA 1984, 81, 1747-1750. Einhorn S, Showe L, Ostlund L, Juliusson G, Rob?rt K-H, Croce C. Influence ofinterferontx on the expression of cellular oncogenes in primary chronic lymphocytic leukemia cells. OncogeneRes (in press). Brenning G, iihre A, Nilsson K. Correlation between in vitro and in vivo sensitivity to human leukocyte interferon in patients with multiple myeloma. &and J Haematol 1985, 35,

543-549.