CELLULAR IMMUNOLOGY 138, 55-63 (199 1)

Identification and Characterization of a T Cell Growth Inhibitory Factor Produced by K562 Erythromyeloid Cells

MARIO DE FELICE,* HEATHER M. BOND,?‘] ROSA PIzzANo,t

MARIA CATERINA TURCO,**t GIULIANA VALERIO$~ ANNALISA LAMBERTI,*

PATRIZIA CARANDENTE GIARRUSSO,? AND SALVATORE VENUTA*,’

*Dipartimento di Medic&a Sperirnentale e Clinica. Facoltd di Medicina, Catanzaro, Ita1.v; tDipartimento di Scienze Biochimiche, and $Dipartimento di Pediatria, II Facoltri di ,bfedicina, Nap&,

Italy

Received April 16, 1991; uccepted June 2, 1991

Cells of the human erythroleukemia cell line K562 constitutively secrete a factor that inhibits human T lymphocyte proliferation induced via CD3/Ti. The factor, termed K-TIF (K562-derived T cell inhibitory factor) is produced in either the presence or absence of fetal calf serum in cultures of K562 cells and can be precipitated by 70% NH,SO,. Gel filtration chromatography on Superose 12 resin by FPLC showed that the inhibitory factor has a molecular weight of approximately 30- 35 kDa. A protein of this size, metabolically labeled with [35S]methionine, specifically bound human peripheral blood mononuclear cells. Chromatofocusing with Mono P by FPLC (pH gradient 7.2-5) indicates that the inhibitory factor has an isoelectric point of 6.0-6.4. 8 1991 Academic

Press. Inc.

INTRODUCTION

The mechanisms by which tumor cells avoid destruction by the host immune system are presently unknown. Immunosuppression of cancer host cells by tumor-derived factors is believed to play a role. Indeed, factors suppressing T cell functions have been found in cancer patients’ serum (1) as well as in tumor tissue extracts (2-4) and medium conditioned by tumor cells in vitro (5- 10). These suppressive factors identified have molecular weights ranging from 7 to 200 kDa. Their biological activities involve for the most part inhibition of interleukin-2 production and suppression of T lym- phocytes (2, 6-8).

In this study a suppressive factor produced by the human erythroleukemia cell line K562 (K562-derived T cell inhibitory factor, K-TIF) is identified and characterized for its biological and biochemical properties.

’ Supported by fellowships from the Commission of European Communities and European Cancer Research Campaign.

* Recipient of a fellowship from Associazione Italiana per la Ricerca sul Cancro. 3 To whom correspondence should be addressed at Dipartimento di Biochimica e Biotecnologie Mediche.

II Facolta di Medicina e Chirurgia. Via S. Pansini 5. 80 13 1 Napoli, Italy.

55

0008-X749/91 $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproductmn in any form resewed.

DE FELICE ET AL. MATERIALS AND METHODS

Cells. K562 cells were grown in RPM1 1640 medium supplemented with 10% heat- inactivated fetal calf serum (FCS, Seromed).

Peripheral blood mononuclear cells (PBMC) were isolated from peripheral blood samples of normal donors by centrifugation through a Ficoll-Hypaque (Sigma Chem- ical Company, St. Louis, MO) density gradient at 400g for 30 min, washed twice with PBS, and resuspended in RPM1 1640 medium supplemented with 10% FCS.

Proliferation assays. These were performed essentially as has been previously de- scribed (11). Briefly, PBMC (106/ml) were incubated in 96-well microtiter plates (Falcon Labware, Becton-Dickinson) in the presence of the anti-CD3 monoclonal antibody (mAb) OKT3 (Ortho Diagnostic System, Milan, Italy) (20 rig/ml) and suppressor factor preparations to be tested (100 &assay), at 37°C in a 5% CO2 atmosphere for 44 hr. Then 50 ~1 of 0.5 &i of [3H]thymidine (specific activity 47 Ci/mmol, Amersham International pcl, Milan, Italy) was added to each well. Following a 4-hr incubation, cultures were harvested on glass fiber strips with a PhD cell harvester (Cambridge Technology, Inc., Cambridge, MA) and [3H]thymidine incorporation was measured in a beta counter. Each assay was performed in duplicate with two different preparations of PBMC. Extracts containing suppressor activity were tested in triplicate.

Preparation of crude suppressive factor. K562 cells were incubated at I X 106/ml in RPM1 1640 medium without FCS for 20 hr. Cells were centrifuged at 400g for 10 min, supematant was harvested and precipitated with 70% NH&SO4 at 4°C. The pre- cipitate was recovered by centrifugation at 12,000g for 30 min, resuspended in phos- phate buffered saline (PBS), and extensively dialyzed against PBS (NH4S04 fraction).

Gelfiltration. The NH4S04 fraction (200 ~1, derived from 200 ml conditioned su- pernatant) was applied to a Superose 12 (HR 1 O/30) column in a fast performance liquid chromatography (FPLC) apparatus (Pharmacia, Uppsala, Sweden) and eluted with PBS at a flow rate of 1 ml/min. Fractions were dialyzed against PBS and tested in duplicate in proliferation assays.

Chromatofocusing. NH&Sod-precipitated material derived from 100 ml of K562 secretion was dissolved in 4 ml of 25 mA4 Tris/HCl, pH 7.1, and dialyzed against 25 rnkf Tris/HCl, pH 7.1. After dialysis remaining insoluble material was removed by centrifugation at 10,OOOg for 10 min and the soluble material applied to a Mono P HR 5/20 column in FPLC apparatus and run in 1: 10 polybuffer 74 (Pharmacia) ad- justed to pH 5.0, at a flow rate of 0.5 ml/min. The column was washed with 25 mM Tris/HCl, pH 7.1 (40 ml), eluted with a pH gradient using 1: 10 polybuffer 74, pH 5.0 (80 ml), and finally washed with 3 M NaCl. Fractions obtained were dialyzed against PBS and tested in proliferation assays.

Metabolic labeling of proteins secreted by K562 cells. K562 cells (20 X 106) were incubated in the absence of FCS in methionine free (MEM) medium supplemented with 20 &i/ml of [35S]methionine (500 mCi/mmol, Amersham International plc) for 20 hr. Cells were then centrifuged and the supematant was precipitated with 70% NH$04 at 4°C. Precipitated material was solubilized in 200 ~1 PBS and applied to a Superose 12 gel filtration column run in PBS. One-milliliter fractions were collected and counted for radioactivity.

Analysis of metabolically labeled proteins from K562 cells and binding to PBMC. Fractions of 35S-metabolically labeled proteins from K562 cells were prepared as de-

A T CELL INHIBITORY FACTOR 57

scribed above. Each fraction (50 ~1) was precipitated with 10% (w/v) trichloroacetic acid (TCA) and analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel elec- trophoresis (PAGE). Samples were reduced by treatment with 5 mM dithiothreitol (DTT). In parallel 200 ~1 of each fraction was incubated with 2 X lo6 PBMC for 2 hr at 37°C with agitation. Cells were washed two times with PBS and solubilized with 1% NP40 10 mM Tris/HCl, pH 8.0, soluble material was precipitated with TCA, and samples were prepared for SDS-PAGE. Gels were fluorographed and exposed to X- ray film for 2-10 days.

RESULTS

Inhibition of Stimulated T Lymphocyte Proliferation by NH4S04 Fractions Obtained jiiom K562 Cell Supernatants

Normal PBMC cells were stimulated with the mitogenic anti-CD3 (monoclonal antibody OKT3). As anti-CD3 mAb stimulates only the T lymphocytes, the suppressor activity measured will affect specifically this cell population.

Preliminary experiments indicated that supernatants of K562 cells, cultured in either the presence or absence of heat-inactivated FCS, could inhibit the proliferation of normal T lymphocytes stimulated in vitro with the mitogenic anti-CD3 mAb OKT3 (not shown). Supernatants from other cell lines, including Hep3B or Hela had no inhibitory effect (not shown). K562 cell supernatants were precipitated with 70% NH4S04, as described under Materials and Methods, and these fractions were tested (after solubilization and dialysis against RPMI) for their inhibitory activity on the proliferation of 12 different preparations of PBMC stimulated with mAb OKT3. Results

TABLE 1

Inhibitory Effect of NH,S04 Precipitated Fractions from K562 Cell Supernatants on the Proliferation of Peripheral Blood Mononuclear Cells

[3H]Thymidine inc. (cpm)

Sample number -pr +pr Percentage of inhibition

I 104205 79330 34 2 107405 59865 44 3 30810 10560 66 4 16660 13735 82 5 67960 26675 61 6 62790 12360 81 7 87930 55240 37 8 73880 37090 50 9 33392 4192 87

10 89185 2135 98 11 31511 2400 92 12 24415 9593 61

Note. K562 cell supematants were precipitated with NH,S04 and the precipitated fractions (pr), resuspended at a 2X concentration with respect to the original supematants, were tested at a I:8 dilution in proliferation assays. Results are means of triplicate determinations; standard deviations were lower than 20%.

58 DE FELICE ET AL.

are illustrated in Table 1. In all cases inhibition was found. There was, however, some variation in the amount of inhibition between different PBMC preparations. In 9 out of 12 PBMC preparations tested (at dilutions of 1:8), [3H]thymidine incorporation in cellular DNA was inhibited by more than 50% (range: 50-98%). In the remaining 3 PBMC preparations, inhibition was 37-44%. An example of the dose dependence curve for the factor is shown in Fig. 1.

Molecular Weight and Isoelectric Point (pIj of the K562 Cell-Derived Suppressive Factor

To determine the molecular weight of K562-derived tumor inhibitory factor, NH&SO4 precipitates from K562 supernatants were subjected to gel filtration chro- matography on Superose 12 using an FPLC system; each fraction was tested for in- hibition of OKT3-stimulated PBMC. Inhibitory activity was recovered in a fraction corresponding to a MW of 30-35 kDa (Fig. 2). There was additionally a small amount of inhibitory activity eluting in the void volume, presumably aggregated material. When K562-derived precipitates were applied to a Mono P column for chromatofo- cusing (pH gradient 7.2-5), the inhibitory activity was recovered at a pl of 6.0-6.4 (Fig. 3).

Binding of K562-Secreted Proteins to PBMC

K562 cells were metabolically labeled with [35S]methionine and the secreted proteins were precipitated with 70% NH4S04. In the same way as for unlabeled suppressor factor preparations the [35S]methionine labeled proteins were fractionated by gel fil- tration (Superose 12). The fractions were processed for SDS-PAGE and the radioactive material is shown in Fig. 4. Fraction 13 corresponds to the peak of suppressor activity at 30-35 kDa. This fraction contains several labeled proteins in this size range. To

40000 1 30000

1

dilution

FIG. 1. Inhibition of human PBMC proliferation. Medium conditioned by K562 cells (see Materials and Methods) was precipitated with 70% NH,SO,, the precipitate was centrifuged, resuspended at a 2X concen- tration (relative to the medium conditioned by K562 cells) in PBS, and extensively dialyzed against PBS. The preparation was tested in a PBMC proliferation assay at different dilutions (1: 1, 1:2, 1:4, 1:8, 1: 16, I: 32). In the absence of inhibitor, [‘Hlthymidine average incorporation in the proliferating cells was 33,000 cpm, as indicated by the line.

A T CELL INHIBITORY FACTOR 59

A

120

100

80

60

40

20

T I/ L

- I

0 1’0 2'0

fraction

T

0.03

0.02

8 cu n 0

0.01

0.00

B 1000000

1

3 100000~ I

10000 ;\.,

10 11 12 13 14 15 fraction

FIG. 2. Fractionation of proteins secreted by K562 cells by Superose 12 gel filtration. K562 cells (2 X IO*/ 200 ml) were cultured in RPM1 medium without FCS for 20 hr. The cells were centrifuged at IOOOg and the soluble proteins precipitated with 70% NH4S04 as described in the methods. This precipitate was solubilized in PBS (200 ~1) and was applied to a Superose I2 (HR 10/30) column using a FPLC apparatus. The column was run in PBS at a flow rate of I ml/min and the void volume was calibrated to be at fraction 6. (A) Fractions were monitored for protein concentration (OD 280) (D) and each fraction was dialyzed against PBS and tested for inhibition of proliferation (Cl). (B) The gel filtration column was calibrated with 200 pg of standard proteins (ribonuclease, 13.7 kDa; ovalbumin, 43 kDa; bovine serum albumin, 68 kDa; transfenin, 80 kDa; aldolase, I58 kDa; and catalase. 232 kDa). The peak of the elution of each protein is shown (0).

determine whether one of these proteins was specifically able to interact with PBMC cells and thus mediate T cell suppressor activity, each of these fractions was incubated with PBMC for 2 hr at 37°C. Cells were washed and subjected to SDS-PAGE and gels fluorographed. Figure 5 shows that in fraction 13 a band of 35 kDa was associated with the PBMC cells; this fraction corresponds to the peak of inhibitory activity obtained in parallel with nonmetabolically labeled extracts, as shown in Fig. 2.

DISCUSSION

A factor produced by K562 erythroleukemia cells (K-TIF) is identified by its ability to inhibit the proliferation of peripheral blood mononuclear cells. Since the proliferation

60 DE FELICE ET AL.

A

80

60

0 10 20 30

fraction

0.06

5% 0 10 20 30 40

fraction

FIG. 3. Fractionation of proteins secreted by K562 cells by Mono P chromatography. K562 cells (lo’/ 100 ml) were cultured in RPM1 medium without FCS for 20 hr. The cells were centrifuged at 1OOOg and the soluble proteins precipitated with 70% NH4S04 as described under Materials and Methods. The precipitated material was dissolved in 4 ml of 25 mM Tris/HCl pH 7.1 and dialyzed against 25 mM Tris/HCl pH 7.1. After dialysis the insoluble material was removed by centrifugation at 10,OOOg for 10 min and the soluble material applied directly to a Mono P column using a FPLC apparatus. The column was washed with 25 mM Tris/HCl, pH 7.1 (40 ml), eluted with a pH gradient using 1: 10 polybuffer 74, pH 5.0 (80 ml), and finally washed with 3 MNaCl starting at fraction 28. (A) Fractions of4 ml each were collected and monitored for protein (OD 280) (w). Each fraction was dialyzed against PBS and tested for inhibition of proliferation (0). (B) The pH of each fraction was measured (0).

was induced via the CD3 antigen the effect must be on the T lymphocyte population. It is, however, possible that the suppressor activity is acting indirectly via another cell population present in PBMC to inhibit T cell proliferation.

The suppressor factor (K-TIF) present in the conditioned medium of K562 cells was effective on all the different PBMC preparations tested; there was, however, some variability in activity between the different batches of PBMC used. This was presumably because of the variability of T cell response to factors controlling T lymphocyte pro- liferation in individual blood samples ( 12). The inhibitory activity was concentration dependent and the effect could be titrated out in each assay. The factor was consti- tutively produced by the K562 cells, and after a 24-hr period in either the presence

A T CELL INHIBITORY FACTOR 61

M.W. 7 8 9 10 11 12 13 14 15 16 17 18 19

97

66

43

31

FIG. 4. Metabolic labeling of proteins secreted by K562 cells. K562 cells (20 X 10h) were metabolically labeled with [35S]methionine (0.2 mCi/lO ml) for 20 hr. in medium containing no methionine or FCS. The cells were centrifuged and the supematant containing secreted protein precipitated at 4°C with 70% NH,S04. After centrifugation, precipitated material was solubilized in 200 ~1 PBS and applied to a Superose 12 gel filtration column run in PBS, as described in Fig. 2. One-milliliter fractions were collected and counted for radioactivity. Fifty microliters of each fraction were precipitated with 10% (w/v) TCA and analyzed by SDS- PAGE with the addition of 5 mM DTT (reducing conditions).

or absence of serum there was no alteration in activity. K-TIF was relatively unstable; after a few days at 4°C or repeated freezing and thawing, there was a considerable loss of activity. When extracts were fast frozen and stored at -80 or -135°C the majority of activity was retained (results not shown).

Fractionation of conditioned medium from K562 cells gave a MW of 30-35 kDa, by gel filtration, and a pl of 6.0-6.4 by chromatofocusing. The biophysical charac- teristics allow a comparison with other such factors produced by neoplastic cells, previously described by ourselves and others. A colon cancer cell line appears to secrete a suppressive factor of 56 kDa (9). Two melanoma cell lines have been found to produce suppressive activities of 100, 7, and 43 kDa, respectively (6, 7). Esophageal cancer (3) and liposarcoma (4) cells have been reported to produce inhibitory activities of 70 and 75 kDa, respectively. A similar or higher (88 kDa) molecular weight has been attributed to suppressive factor(s) derived from T cell leukemias ( 13, 14). Finally, we described previously a 85-kDa transferrin-like protein, produced by human cells derived from a Sezary syndrome, with inhibitory activity on T lymphocyte proliferation ( IO). Additionally the pl of K-TIF (6.0-6.4) is different from those reported for other inhibitory proteins (4, 6, 9, 12). To our knowledge, 30- to 35-kDa suppressor factors with pl of 6.0-6.4 have not so far been identified.

The active inhibitory fraction from the gel filtration column was found to contain a 35-kDa polypeptide having the specific ability to be associated with PBMC cells after an incubation at 37°C for 2 hr. This polypeptide is therefore a probable candidate for the K-TIF activity. Binding experiments at 4°C did not show any specific interaction for this polypeptide (not shown): this could, however, be due to a reduced affinity at a lower temperature. From this approach it would be predicted that a direct interaction

62 DE FELICE ET AL.

M.W. 76 9 10 11 12 13 14 15 16 17 18 19

200

43

31

FIG. 5. Analysis of metabolically labeled proteins from K562 cells and binding to PBMC. Fractions of 35S-metabolically labeled proteins from K562 cells were obtained after Superose 12 gel filtration as described above. Each fraction (200 r.d) was incubated with 2 X lo6 PBMC for 2 hr at 37°C with shaking. Cells were then washed twice with PBS at 4°C and resuspended in 10 mM Tris/HCl, pH 8.0, soluble material was precipitated with 10% TCA, and samples were prepared for SDS-PAGE with 5 mM DTT. Gels were fluo- rographed and exposed to X-ray film for 10 days.

of this 35kDa polypeptide with the cell surface results in suppression of T cell growth. In conclusion, the results presented show that K562 cell-derived inhibitory activity

differs from all other suppressive factors derived from tumor tissues, peripheral blood leukemic cells, and cell lines. Whether this factor is produced also by other erythroid and/or myeloid cell lines and by leukemic cells from peripheral blood of patients remains to be determined. Molecular cloning of the cDNA for K-TIF, which is presently underway, will help investigate this issue and dissect the mechanism of action of this leukemia-derived immunomodulator.

ACKNOWLEDGMENTS

The authors are grateful to Ms. Rita Bisogni and Mr. Attilio Visconti for their excellent technical help and to Ms. Cristina Maresca for her excellent secretarial assistance. This work was supported by funds of P.F. “Biotecnologie e Biostrumentazione” of the Italian C.N.R.

REFERENCES

1. Hellstrom, K. E., and Hellstrom, I., In “Advances in Immunology” (F. Dixon and H. Kunkel, Eds.), Vol. 18, p. 209. Academic Press, New York, 1974.

2. Remacle-Bonnet, M. M., Pommier, G. J., Kaplanski, S., Rance, R. J., and Depieds, R. C., J. Zmmunol. 117, 1145, 1976.

3. Mohagheghpour, N., Parhami, B., Dowiatshahi, K., Kadijehnouri, D., Elder, J. H., and Chisari, F. V., J. Immunol. 122, 1350, 1979.

4. Roth, J. A., Osborne, B. A., and Ames, R. S., J. Immunol. 130, 303, 1983.

5. Whitehead, J. S., and Kim, Y. S., Cancer Res. 40, 29, 1980. 6. Werkmeister, J., Zaunders, J., McCarthy, W., and Hersey, P., Clin. Exp. Immunol. 41, 487, 1980.

7. Hersey, P., Bindon, C., Czerniecki, M., Spurling, A., Wass, J., and McCarthy, W. H., J. Immunol. 132,

2837, 1983.

A T CELL INHIBITORY FACTOR 63

8. Fontana, A., Hengartner, H., de Tribolet, N., and Weber, E., J. Irnrnunol. 132, 1837, 1984. 9. Ebert, E. C., Roberts, A. I., O’Connell, S. M., Robertson, F. M., and Nagase, H., J. Irnmunol. 138,

2161, 1987. IO. Morrone, CL, Corbo, L., Turco, M. C., Pizzano, R., De Felice, M., Bridges, S., and Venuta, S., Cancer

Res. 48, 3425, 1988. I I. De Felice, M., Turco, M. C., Carandente Giarrusso, P., Corbo, L., Pizzano, R., Martinelli, V., Ferrone,

S., and Venuta, S., J. Immunol. 139, 2683, 1987. 12. Carandente Giarrusso, P., Turco, M. C., Corbo. L., Maio, M.. Alfinito, F., Scala G., Zappacosta. S.,

and Venuta, S.. Tissue Antigens 31, 59, 1988. 13. Shirakawa, F.. Tanaka, Y., Oda, S., Chiba, S., Suzuki, H., Eto, S., and Yamashita, U., Chncer Rex 46,

4458, 1986. 14. Santoli, D., Tweardy. J.. Ferrero, D., Kreider, B. L., and Rovera, G., J. hp. Med. 163, 18, 1986.