- Email: [email protected]
Killing of human tumor cells in culture with adriamycin conjugates of human transferrin
CiINICAL
IMIvKJNOLOGY
AND
IMMLJNOPATHOLOGY
32, l-11 (1984)
Milling of Human Tumor Cells in Culture with Adriamyci Conjugates of Human Transferrin CHANG-JING G. YEH AND W. PAGE FAULK INSERM
U210, Laboratoire
d’immunologie, Faculte’ de Medecine, Avenue de Vallombrose, Nice-Ce’dex, France
06034
Receptors for human transferrin (Tff) in high density are found on reticulocytes and syncytiotrophoblast, but most unstimulated, normal adult cells do not bind Tif. In contrast, leukemia and breast adenocarcinoma cells have been shown to manifest Trf receptors, raising the possibility that these receptors might be employed to bind cytotoxic Trf conjugates. Trf was conjugated with adriamycin (Adr) and it was shown that the conjugates are bound by Trf receptors on plasma membranes of Daudi and HL-60 cells, following which Adr is identified in the nuclei of these cells. The biological effect of Adr is quantitated by the inhibition of tridiated thymidine uptake, and subsequent cell death is measured by trypan blue exclusion. The killing correlates directly with both the time of exposure and the amount of conjugate employed. These results suggest that such cytotoxic Tlf conjugates hold promise for selective in vivo killing of some malignant cells.
INTRODUCTION
In addition to their well-known location on reticulocytes, we and other investigators have identified transferrin (Ti-f) receptors on syncytiotrophoblast (1, 2), human cells maintained in culture (3, 4), breast adenocarcinoma tissues (5, 6), and peripheral blood mononuclear cells (PBM) from patients with leukemia (79), while the vast majority of normal PBM do not bind Tl-f (1, 3, 4, 7-9). Biochemical analysis of Tif receptors has also been done with the use of monoclonal antibodies against the receptor (10, ll), and the expression of this receptor has heen assigned to chromosome 3 (12, 13). Apart from iron transport, its biological functions still remain undefined. Nonetheless, evidence has accumulated that Trf, following binding to receptors on plasma membranes, is internalized (14-17). We have proposed that this mechanism might facilitate the transport of Tsf into receptor-bearing cells (5), putting forward the possibility of preparing conjugates of Ti-f with antitumor drugs to allow a piggyback system whereby drugs could be transported into the cells. To test this hypothesis, we have linked Trf with adriamycin (Adr) and found that such conjugates were detrimental to receptor-bearing cells in vitro while having much less effect on receptor-negative normal cells. Data in support of this concept are the subject of this report. MATERIALS
AND METHODS
Cells. HL-60 (acute promyelocytic leukemia) and Daudi (Burkitt’s lymphama) ceils (18), were grown in suspension at 37°C in RPMI-1640 medium with L-glutamine, 25 rr&Z Hepes (Gibco Europe, Glasgow, Scotland), 50 pg/ml gentamicin 1
0090-1229/84 $1 SO Copyright Ail rights
0 1984 by Academic Press, Inc. of reproduction in any form reserved.
2
YEH
AND
FAULK
(Flow Laboratories Ltd., Ayrshire, Scotland), and supplemented with 10% fetal calf serum (Sera-Lab Ltd., Crawley Down, Sussex, England), this combination being hereafter referred to as complete medium. Human venous blood was collected into acid citrate dextrose solution, and PBM were separated from venous blood samples with the use of Isolymph (Gallard-Schlesinger Chemical Mfg. Corp., Carle Place, N.Y.). Viability was determined by trypan blue exclusion. Preparation of conjugates. To a solution containing 10 mg of either human Trf (99% iron free, from Behringwerke AG, Marburg, W. Germany) or human cernloplasmin (Cer) (prepared by Dr. M. Steinbuch of the French National Blood Transfusion Center, Orsay, according to reference 19>, and 3 mg of Adr hydrochloride (Farmitalia Co., Milan, Italy) in 1 ml of 0.1 M PBS, pH 7.0, was added dropwise 0.5 ml of an aqueous solution of 0.25% glutaraldehyde (BDH Chemicals Ltd., Poole, England) at room temperature (RT) with gentle mixing (20). After 2 hr incubation at RT in dark, 0.5 ml of 1 M ethanolamine (Sigma London Chemical Co. Ltd.), pH 7.4, was added and incubated at 4°C overnight. The mixture was then centrifuged at 1OOOg for 15 min and the supernatant was collected and chromatographed through a column (2 x 20 cm) of Sepharose CL-6B (Pharmacia Fine Chemicals, Uppsala, Sweden) equilibrated in 0.15 M PBS, pH 7.2. Protein and Adr were identified as they emerged from the column by spectrophotometric readings taken at OD,,, and OD,,, for Trf or Cer and Adr, respectively (21). Spectrophotometrically defined 1.2-ml fractions were pooled and sterilized by irradiation in a gamma cell 1000 irradiator (Atomic Energy of Canada) and stored at 4°C in dark until use. Zmmunocytology andfluorescence localization of Trf and Adr. Cells (1 x 106) were reacted with 100 p,l of Adr, Trf, or Trf-Adr conjugate for 20 min at 4°C or for 3 hr at 37°C washed twice in Hank’s balanced salt solution (HBSS, Gibco Europe) at 4°C and reacted with 100 ~1 of a 1:20 dilution of a fluorescein isothiocyanate (FITC) conjugate of goat anti-human Trf (Atlantic Antibodies, Maine). This antibody did not react with Daudi or HL-60 cells prior to their exposure to human Tl-f. Following incubation with anti-T& the cells were washed three times in HBSS at 4°C suspended in 50% glycerol/PBS, and mounted on glass slides to be studied by epi-illumination with a Zeiss universal microscope fitted with an HBO 50 mercury arc lamp. This microscope was equipped with three different optical systems (22) for the analysis of Trf and Adr: (a) the FITC labeled anti-Trf could be identified by using a blue interference filter (BP 455-490) with a FT 510 chromatic beam splitter and a BP 520-560 barrier filter; (b) Adr was identified by using a green interference filter (BP546/7) with a FT 580 chromatic beam splitter and a LP 590 barrier filter; and (c) anti-Trf and Adr could be simultaneously identified by using a BP455-490 interference filter, a FT 510 chromatic beam splitter, and a LP520 barrier filter. Measurements of cellular proliferation and viability. For a direct inhibition assay, cells (0.5 x 106) were incubated with 100 p,l of sterile Tif-Adr or with Cer-Adr conjugates containing 1 pg of Adr, or with complete medium at 37°C in 5% CO,/air for 3 hr, following which they were washed twice in complete medium for 7 min per wash at 4°C and centrifuged at 400g. For a competitive inhibition assay, cells (0.5 x 106) were incubated with or without 100 ~1 of Trf (1
KILLING
a.
Native
TUMORS
WITH CYTOTOXIC
Trf
c. Trf-Adr
*--A---+-B’
3
TRANSFERRIN
con,ugate
r-c-3
10
20
30
40
FIG. I. Chromatographic properties of Trf, Trf-Adr conjugates, and Adr on columns of Sepharose CL-f%%. (a) Native TrE; (b) Trf control; (c) Tif-Adr conjugate. - OD?,,; ---- OD,9,.
mgiml) at 4°C for 30 min, followed by incubation with Trf-Adr, Cer-Adr, or PBS at 4°C for 30 min, they were then washed twice at 4°C and further incubated at 37°C for 2 hr. For the quantitation of cellular proliferation, 300 ~1 of complete medium was added to the washed cell pellet, and the cells were resuspended and distributed in triplicate as 100~l.~l aliquots in flat-bottom microtiter plates (Flow Laboratories), each well of which was preloaded with 100 ~1 of complete medium and 25 ~1 of 2 @Yi of [3H]thymidine (sp act., 2 Ci/mmol, Radiochemical Center, Amersham, Bucks., England). Plates were then incubated at 37°C in 5% C&/air for 16 hr, and the cells were harvested on glass fiber filters in a Mash II multiple harvester (Microbiological Associates, Walkersville, Md.). The amount of i3H]thymidine incorporated into DNA was measured in a LKB 1216 Rackbeta II liquid scintillation counter and expressed as counts per minute (cpm). Quantitative Ifleasures of the viability and number of cells were also done in all experiments.RESULTS Chromatographic Properties of Trfi Trf-Adr
Conjugates, and Adr
In preparing the conjugates, we found that native Trf (1 ml of 5 mglml in P consistently emerged from a column of Sepharose CL-6B between fractions 22 and 30, peaking at fractions 26 (Fig. la). However, when Trf, which had been carried through the conjugation procedure in the absence of Adr, was chromatographed on the same column, a more complex pattern was obtained (Fig. lb). The macromolecular peak (A in Fig. lb) was precipitated by anti-Trf antibody in Buchterlony and immunoelectrophoresis gels, as did the samples which emerged
4
YEH
AND
FAULK
FIG. 2. Immunological identification of Tti on cells and fluorescence localization of Adr in nuclei. Note uniformly speckled pattern of membrane (green) fluorescence produced by FITC anti-Trf (a), nuclear (red) fluorescence produced by Adr (b), and cells with both membrane and nuclear fluorescence following incubation with Tif-Adr conjugates (c). (X 625).
at the same place as native Trf, i.e., between fraction numbers 20-26 (B in Fig. lb), but the late appearing peak (C in Fig. lb) was not precipitated by anti-Tlf in gels. For the purposes of this study, “A” was considered to be Trf aggregates, “B” was chromatographically and antigenically native Trf, and “C” was Trf fragments. When Trf- Adr conjugates were passed through the column, the OD,,, tracing was similar to the controls, but reading at ODdg, revealed Adr in both peaks A and B (Fig. lc), whereas nothing was detected at OD,,, in the absence of Adr. Inasmuch as peak B was antigenically native Tti and its spectrophotometric characteristic indicated the presence of Adr, it was used as Tif-Adr conjugates in the following experiments. It should be noted that Adr has a molecular weight of 579.98 (23), and that it does not chromatograph within the molecular weight range of Sepharose CL-6B.
KILLING
TUMORS
WITH
FIG.
CYTOTOXIC
TRANSFERRIN
2-Continued.
Immunological IdentiYcation of Trf on Plasma Membranes and Fluorescence Localization of Adr in Nuclei
Trf, Adr, and Trf-Adr conjugates were investigated for their ability to be bound by plasma membrane receptors and to intercalate with nuclear DNA (24). This was done by using two different manipulations: the first being the incubation of Trf receptor-positive cells with the above reagents for 20 min at 4°C and the second being incubation of cells with the same reagents for 3 hr at 37°C. For these experiments, Daudi and HL-60 cells were used, since we have previously reported that their plasma membranes have receptors for Tif (7, I@, and PB were used as Tif receptor-negative cells (3, 4, 7-9). When HL-60 or Daudi cells were incubated with Trf which had been carried through the conjugation and chromatography in the absence of Adr, or with Ti-f-Adr conjugates for 20 min at
6
YEH
AND
FAULK
4°C a uniformly speckled pattern of fluorescence could be identified on their plasma membranes following incubation with FITC-labeled antibodies to human Trf (Fig. 2a). In contrast, only 8% of PBM were found to bind Trf or Trf-Adr conjugates, and none of the PBM, HL-60, or Daudi cells reacted with FITC antiTrf after being incubated with Adr alone (Table 1). However, incubation of PBM, HL-60, or Daudi cells with free Adr for 20 min at 4°C resulted in a red fluorescence of their nuclei (Fig. 2b), a result which was not obtained by incubation of any of the target cells with Trf or Trf-Adr conjugates for 20 min at 4°C. When these experiments were repeated with the use of the same cells and the same reagents, but with the incubation time changed to 3 hr and the temperature changed to 37°C the results were different. In the first instance, the pattern of immunofluorescence following incubation of HL-60 or Daudi cells with Trf or Trf-Adr conjugates was no longer homogeneous, but appeared to be clustered into islands of fluorescent patches on plasma membranes (Fig. 2~). The most striking difference found with these changed conditions of time and temperature for incubation was that with the Tif-Adr conjugates, for Adr could now be identified in the nucleus (Fig. 2~).
Effects of Trf-Adr
Conjugates
on [‘Hjthymidine
Uptake and Viability
of Cells
It is thought that Adr acts by inserting itself between base pairs of DNA (25), resulting in an inhibition of DNA synthesis and eventual cell death, hence we chose the inhibition of [3H]thymidine uptake as well as trypan blue exclusion as measures of cellular proliferation and viability, respectively. Incubation of TrfAdr conjugates for 3 hr at 37°C with Daudi and HL-60 cells followed by two washes and a 16-hr culture period in tritiated thymidine had the striking effect of diminishing both proliferation and viability, whereas minimal effects were observed for both thymidine incorporation and viability when PBM were tested under the same conditions (Table 2). The Trf control manifested moderate suppression of Daudi and HL-60 cells and augmentation of PBM proliferation, while Tif-Adr conjugates consistently recorded more than 90% suppression of both cell lines in 30 experiments. Regarding the death of PBM, it should be pointed out that, unlike Daudi and HL-60, very few PBM were found to contain Adr in their nuclei, indicating that dead cells resulted as a consequence of manipulation in vitro. In order to provide additional evidence for the specific inhibition of cellular proliferation by Tif-Adr conjugates, Cer-Adr conjugates were used as a protein control for Trf. The choice of Cer was predicated by the observation that other proteins such as human albumin, Gc, and alpha fetal protein are immunologically cross-reactive with unfolded Tif but not with Cer (26, 27). Although Cer-Adr conjugates were found to kill target cells in vitro, when Trf-Adr or Cer-Adr conjugates were allowed to displace Trf, Trf-Adr conjugates were almost twice as efficient as Cer-Adr conjugates (Table 3). Native Ti-f alone in the growth media
KILLING
TUMORS
WITH CYTOTOXIC TABLE
PLASMA
MEMBRANE
Reagents
TRANSFERRIN
7
1
AND NUCLEAR REACTIVITY OF Trf, Adr, AND Trf-Adr CONJUGATES CELLS AT DIFFERENT TEMPERATURES AND TIMES OF INCUBATION
Temperature (“C)
Tlf A& Trf-Adr Trf Adr Trf- Adr
Membrane fluorescence for Tif
Time
4 4 4 37 37 37
20 min 20 min 20 min 3 hr 3 hr 3 hr
a Percentage of cells fluorescing/intensity b Percentage of positive cells.
70-SO/3 0 70-8013 70-8013 0 70-SO/3
WITH
HL-60
Nuclear fluorescence for Adr
+n
0 1OOb 0 0 100 70-80
+ + +
of fluorescence (Ref. 18).
could not have been solely responsible for these results, since test and control experiments contained the same amount of Trf (Table 3). Dose and Time Variables
in Cell Killing
by Trf-Adr
Conjugates
Ti-f-Adr conjugates were serially diluted to test the endpoint of their ability to inhibit the uptake of [3H]thymidine. The results of these experiments showed that the proliferation of HL-60 cells was directly associated with the amount of TrfAdr conjugates used in the initial 3-hr preincubation step of the assay (Fig. 3). The percentage inhibition of cellular replication was rapid and achieved a plateau at 30 min preincubation (Fig. 4), although the in vitro assay for inhibition of cellular proliferation employed a preincubation step that was studied by using a xed amount of conjugates with varied periods of time.
TABLE 2 EFFECTS OF
Trf-Adr
ON [3H]T~~~~~~~~
UPTAKE
AND VIABILITY
[3H]thymidine uptake (cpm) Cellsb Daudi HE-60 PBM
OF CELLS
Viability (W
Control
Trf control
Trf- Adr
Control
Trf control
Trf- Adr
155,511 +- 6,793” 199,918 r 7,489 316 r 87
72,339 2 3,159 32,058 k 1,899 672 IL 215
14,850 i 812 3,515 i 124 98 r 25
89 92 94
73 88 87
63 56 86
a Mean 2 SEM from data of triplicate cultures obtained from 30 experiments. b Cells (0.5 x 106) were incubated with growth medium (control), Trf control (B in Fig. lb), or TrfAdr (B in Fig. lc) for 3 hr at 37”C, washed, and incubated with [3H]thymidine for 16 hr at 37”C, after which the uptake of [3H]thymidine and viability were determined.
8
YEH
AND
TABLE COMPETITIVE
INHIBITION
OF
Trf-Adr UPTAKE
FAULM
3
AND Cer-Adr WITH Trf ON [3H]T~~~~~~~~ OF DAUDI CELLS
Condition Reagent Tif- Adr Cer-Adr Control
No Preincubation
Preincubation
87,000 i 5,857 80,687 f 2,025 145,987 f 6,266
with Trf
48,624 + 3,724 85,563 i 5,677 122,467 ? 7,625
Note. A composite of results from two experiments in which cells were preincubated with or without 100 ~1 of Trf for 30 min at 4”C, followed by incubation with Tif-Adr. Cer-Adr, or PBS for 30 min at 4”C, washed twice at 4”C, and then further incubated for 2 hr at 37°C. The measurement of incorporation of [3H]thymidine was as detailed under Materials and Methods.
DISCUSSION
It has been proposed that some cancer chemotherapeutic agents might be surreptitiously transported into cells as conjugates of Tif through the mediation of Trf receptors (5), and the findings in this report support this idea by showing a temperature-dependent internalization of Ti-f-Adr complexes into the cytoplasm and the subsequent movement of Adr into the nuclei of Daudi and HL-60 cells. One explanation of these results was that TI-f-Adr conjugates were bound by Trf receptors on the cells, and following the 3-hr incubation at 37”C, they were endocytosed into so-called receptosomes (28). In consequence, the covalent bond between Tif and Adr in the conjugates was cleaved in the acidic cytoplasmic compartment (17) to liberate free Adr to diffuse into and be bound by nuclear DNA. The observed cytotoxic effect was directly associated with the amount of conjugate used, as shown by dose-response experiments, and this effect achieved a plateau following an incubation period of 30 min with the conjugate.
FIG. 3. Dose-response HL-60 cells.
of inhibition
of Trf-Adr
conjugates on the uptake of i3H]thymidine
by
KILLING
TUMORS
WITH CYTOTOXIC
9
TRANSFERRIN
20 i’ ’ 1;
t----
,
10
30
FIG. 4. Time course of inhibition cells.
60
of Trf-Adr
,
90 Pre-incubation
120
1
150
180
conjugate on the uptake of [3H]thymidine on Daudi
It has also been proposed that retinoic acid, which induces the differentiation of HL-60 cells, offers another approach to the treatment of some human leukemias (29). Indeed, following induction by retinoic acid or dimethyl sulfoxide, we have shown that the morphological and functional maturation of HL-60, but not of Daudi cells, was accompanied by a loss of Tif receptors (18), thus alterations in both cell types following exposure to Trf-Adr conjugates in the present experiments could not be due to a mechanism similar to that observed with either retinoic acid or dimethyl sulfoxide. Interestingly, ricin conjugates of monoclonal antibodies to Tif receptors or the antibodies alone have also been reported to inhibit the growth of human tumor cells in nude mice (30), again indicating that Trf receptors are present and are useful in studies of the biology of Trf receptors. Conjugates of Tif with Adr would not be expected to generate clinical problems, for Trf is a normal constituent of plasma, and the literature contains many reports of Adr chemotherapy without immunopathological sequelae, suggesting that the use of Trf-Adr conjugates would probably not cause untoward immunological reactions in patients. Furthermore, when leukemic cells were incubated with TsfAdr conjugates in the presence of normal human plasma, the percentage of inhibition remained unchanged (unpublished data), suggesting that the Trfreceptors are available to bind cytotoxic Ti-f from the blood circulation. It is possible that the Trf receptor is a family of structurally related but antigenitally and functionally distinct molecules much as was suspected as a result of studies of breast adenocarcinoma cells (5). This is suggested by observations that some but not all monoclonal antibodies to Trf receptors recognize receptor in human skin (31, 32). Thus, physiological receptors for Trf on normal cells may be irrelevent in our proposed regimen. There are two additional usages of Trf and Trf receptors which could be exploited: the first being that radiolabeled Trf might be used for the detection of primary or metastatic tumors as is now done by using
10
YEH AND FAULK
Gallium (33) or isotopically labeled antibodies to carcinoembryonic antigen (34); and second, the possibility of preparing a battery of conjugates of Trf with other chemotherapeutic agents to assess the relative efficiency of each to kill bone marrow cells in vitro from newly diagnosed leukemic patients as an adjunct in deciding what agent should be employed for therapy. The latter possibility of drug tailoring to fit the patient is supported in part by unpublished data from our laboratory showing that PBM and bone marrow cells from patients with acute myelogenous leukemia were selectively killed in vitro by Trf-Adr conjugates in a fashion similar to that described in this report. The results of these experiments form the basis for a subsequent report. ACKNOWLEDGMENTS This research was supported in part by the British Medical Research Council, INSERM Unit 210, the East Grinstead Research Trust, and the Human Embryology and Pregnancy Foundation. We thank Dr. M. Steinbuch for her gift of human ceruloplasmin.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
Fauik, W. P., and Galbraith, G. M. P., Proc. Roy. Sot. Lorzdon B 204, 83, 1979. Wada, H. D., Hass, P. E., and Sussman, H. H., J. Biol. Chem. 2.54, 12629, 1979. Larrick, .I. W., and Cresswell, P., Biochim. Biophys. Acta 583, 483, 1979. Galbraith, G. M. P., Galbraith, R. M., and Faulk, W. P, Cell. Immunol. 49, 215, 1980. Faulk, W. P., Hsi, B.-L., and Stevens, P. J., Lancet 2, 390, 1980. Shindelman, J. E., Qrtmyer, A. E., and Sussman, H. H., Int. J. Cancer 27, 329, 1981. Yeh, C. G., Hsi, B.-L., and Faulk, W. I?, In “Aspects of Developmental and Comparative Immunology” (J. B. Solomon, Ed.), 361, Pergamon, Oxford, 1980. Larrick, J. W., and Logue, G., Lancet 2, 862, 1980. Yeh, C. G., Taylor, C. G., and Faulk, W. P., VOX Sang&us, in press. Gmary, M. B., and Trowbridge, I. S., J. Biol. Chem. 256, 12888, 1981. Schneider, C., Sutherland, R., Newman, R. A., and Greaves, M. F., J. Biol. Chem. 2.57, 8516, 1982. Enns, C. A., Suomalainen, H. A., Gebhardt, J. E., Schroder, J., and Sussman, H. H., Proc. Natl. Acad. Sci. USA 79, 3241, 1982. Goodfellow, l? N., Banting, G., Sutherland, D. R., Greaves, M. F., Solomon, E., and Povey, S., Somat. Cell Gent. 8, 197, 1982. Sullivan, A., Grasso, J. A., and Weintraub, L. R., Blood 47, 133, 1976. Hemmaplardh, D., and Morgan, E. H., Brit. J. Haematol. 36, 85, 1977. Karin, M., and Mintz, B., J. Biof. Chem. 256, 3245, 1981. van Renswoude, J., Bridges, K. R., Harford, J. B., and Klausner, R. D., Proc. Natl. Acad. Sci. USA 79, 6186, 1982. Yeh, C. G., Papamichail, M., and Faulk, W. P., Exp. Cell Res. 138, 429, 1982. Pegaudier, Z., Andran, R., and Steinbuch, M., Clin. Chem. Acta 30, 387, 1970. Van Vunakis, H., Langone, J. H., Riceberg, L. J., and Levine, L., Cancer Res. 34, 2546, 1974. Hurwitz, E., Levy, R., Maron, R., Wilchek, M., Amon, R., and Sela, M. Cancer Res. 35, 1175, 1975. Yeh, C. G., Hsi, B.-L., and Faulk, W. P., J. Immunol. Methods 43, 269, 1981. Arcamone, E, Cassinelli, G., Francheschi, G., Pence, S., Pol, C., Redaelli, S., and Selva, A., In “international Symposium on Adriamycin” p. 9, Springer-Verlag, Berlin/Heidelberg/New York, 1972. Pigram, W. J., Fuller, W., and Hamilton, L. D., Nature (London) 217, 235, 1972. Zunino, E, Gambetta, R., DiMarco, A., and Zaccara, A., Biochim. Biophys. Acta 277,489, 1972. Pekkala-Flagan, A., and Ruosiahti, E., J. Immunol. 128, 1163, 1982.
KILLING
TUMORS
WITH CYTOTOXIC
TRANSFERRIN
If
27. Metz-Boutique, M. H., Jolles, J., Mazurier, J., Spik, G.; Mantreuil, J., and Jolles, P., FEBS Letf. 132, 239, 1981. 28. Willingham, M. C., and Pastan, I., Cell 21, 67, 1980. 29. Breitman, T. R., Selonick, S. Z., and Collins, S. J., Proc. Natl. Acad. Sci. USA 77, 2936, 1980. 30. Trowbridge, I. S., and Domingo, D. L., Nature (London) 294, 171, 1981. 31. Wells, M., Yeh, C. G., Hsi, B.-L., and Faulk, W. P., J. Clin. Pathol., in press. 32. Gatter, K. C., Brown, G., Trowbridge, I. S., Woolston, R. E., and Mason, 0. Y., J. Clin. Pathoi. 36, 539, 1983. 33. Silberstein, E. B, C/in. Nucl. Med. 6, 63, 1981. 34. Mach, J. P., Oncodev. Biol. Med. 1, 49, 1980. Received July 6, 1983; accepted with revision January 12, 1984.
IMIvKJNOLOGY
AND
IMMLJNOPATHOLOGY
32, l-11 (1984)
Milling of Human Tumor Cells in Culture with Adriamyci Conjugates of Human Transferrin CHANG-JING G. YEH AND W. PAGE FAULK INSERM
U210, Laboratoire
d’immunologie, Faculte’ de Medecine, Avenue de Vallombrose, Nice-Ce’dex, France
06034
Receptors for human transferrin (Tff) in high density are found on reticulocytes and syncytiotrophoblast, but most unstimulated, normal adult cells do not bind Tif. In contrast, leukemia and breast adenocarcinoma cells have been shown to manifest Trf receptors, raising the possibility that these receptors might be employed to bind cytotoxic Trf conjugates. Trf was conjugated with adriamycin (Adr) and it was shown that the conjugates are bound by Trf receptors on plasma membranes of Daudi and HL-60 cells, following which Adr is identified in the nuclei of these cells. The biological effect of Adr is quantitated by the inhibition of tridiated thymidine uptake, and subsequent cell death is measured by trypan blue exclusion. The killing correlates directly with both the time of exposure and the amount of conjugate employed. These results suggest that such cytotoxic Tlf conjugates hold promise for selective in vivo killing of some malignant cells.
INTRODUCTION
In addition to their well-known location on reticulocytes, we and other investigators have identified transferrin (Ti-f) receptors on syncytiotrophoblast (1, 2), human cells maintained in culture (3, 4), breast adenocarcinoma tissues (5, 6), and peripheral blood mononuclear cells (PBM) from patients with leukemia (79), while the vast majority of normal PBM do not bind Tl-f (1, 3, 4, 7-9). Biochemical analysis of Tif receptors has also been done with the use of monoclonal antibodies against the receptor (10, ll), and the expression of this receptor has heen assigned to chromosome 3 (12, 13). Apart from iron transport, its biological functions still remain undefined. Nonetheless, evidence has accumulated that Trf, following binding to receptors on plasma membranes, is internalized (14-17). We have proposed that this mechanism might facilitate the transport of Tsf into receptor-bearing cells (5), putting forward the possibility of preparing conjugates of Ti-f with antitumor drugs to allow a piggyback system whereby drugs could be transported into the cells. To test this hypothesis, we have linked Trf with adriamycin (Adr) and found that such conjugates were detrimental to receptor-bearing cells in vitro while having much less effect on receptor-negative normal cells. Data in support of this concept are the subject of this report. MATERIALS
AND METHODS
Cells. HL-60 (acute promyelocytic leukemia) and Daudi (Burkitt’s lymphama) ceils (18), were grown in suspension at 37°C in RPMI-1640 medium with L-glutamine, 25 rr&Z Hepes (Gibco Europe, Glasgow, Scotland), 50 pg/ml gentamicin 1
0090-1229/84 $1 SO Copyright Ail rights
0 1984 by Academic Press, Inc. of reproduction in any form reserved.
2
YEH
AND
FAULK
(Flow Laboratories Ltd., Ayrshire, Scotland), and supplemented with 10% fetal calf serum (Sera-Lab Ltd., Crawley Down, Sussex, England), this combination being hereafter referred to as complete medium. Human venous blood was collected into acid citrate dextrose solution, and PBM were separated from venous blood samples with the use of Isolymph (Gallard-Schlesinger Chemical Mfg. Corp., Carle Place, N.Y.). Viability was determined by trypan blue exclusion. Preparation of conjugates. To a solution containing 10 mg of either human Trf (99% iron free, from Behringwerke AG, Marburg, W. Germany) or human cernloplasmin (Cer) (prepared by Dr. M. Steinbuch of the French National Blood Transfusion Center, Orsay, according to reference 19>, and 3 mg of Adr hydrochloride (Farmitalia Co., Milan, Italy) in 1 ml of 0.1 M PBS, pH 7.0, was added dropwise 0.5 ml of an aqueous solution of 0.25% glutaraldehyde (BDH Chemicals Ltd., Poole, England) at room temperature (RT) with gentle mixing (20). After 2 hr incubation at RT in dark, 0.5 ml of 1 M ethanolamine (Sigma London Chemical Co. Ltd.), pH 7.4, was added and incubated at 4°C overnight. The mixture was then centrifuged at 1OOOg for 15 min and the supernatant was collected and chromatographed through a column (2 x 20 cm) of Sepharose CL-6B (Pharmacia Fine Chemicals, Uppsala, Sweden) equilibrated in 0.15 M PBS, pH 7.2. Protein and Adr were identified as they emerged from the column by spectrophotometric readings taken at OD,,, and OD,,, for Trf or Cer and Adr, respectively (21). Spectrophotometrically defined 1.2-ml fractions were pooled and sterilized by irradiation in a gamma cell 1000 irradiator (Atomic Energy of Canada) and stored at 4°C in dark until use. Zmmunocytology andfluorescence localization of Trf and Adr. Cells (1 x 106) were reacted with 100 p,l of Adr, Trf, or Trf-Adr conjugate for 20 min at 4°C or for 3 hr at 37°C washed twice in Hank’s balanced salt solution (HBSS, Gibco Europe) at 4°C and reacted with 100 ~1 of a 1:20 dilution of a fluorescein isothiocyanate (FITC) conjugate of goat anti-human Trf (Atlantic Antibodies, Maine). This antibody did not react with Daudi or HL-60 cells prior to their exposure to human Tl-f. Following incubation with anti-T& the cells were washed three times in HBSS at 4°C suspended in 50% glycerol/PBS, and mounted on glass slides to be studied by epi-illumination with a Zeiss universal microscope fitted with an HBO 50 mercury arc lamp. This microscope was equipped with three different optical systems (22) for the analysis of Trf and Adr: (a) the FITC labeled anti-Trf could be identified by using a blue interference filter (BP 455-490) with a FT 510 chromatic beam splitter and a BP 520-560 barrier filter; (b) Adr was identified by using a green interference filter (BP546/7) with a FT 580 chromatic beam splitter and a LP 590 barrier filter; and (c) anti-Trf and Adr could be simultaneously identified by using a BP455-490 interference filter, a FT 510 chromatic beam splitter, and a LP520 barrier filter. Measurements of cellular proliferation and viability. For a direct inhibition assay, cells (0.5 x 106) were incubated with 100 p,l of sterile Tif-Adr or with Cer-Adr conjugates containing 1 pg of Adr, or with complete medium at 37°C in 5% CO,/air for 3 hr, following which they were washed twice in complete medium for 7 min per wash at 4°C and centrifuged at 400g. For a competitive inhibition assay, cells (0.5 x 106) were incubated with or without 100 ~1 of Trf (1
KILLING
a.
Native
TUMORS
WITH CYTOTOXIC
Trf
c. Trf-Adr
*--A---+-B’
3
TRANSFERRIN
con,ugate
r-c-3
10
20
30
40
FIG. I. Chromatographic properties of Trf, Trf-Adr conjugates, and Adr on columns of Sepharose CL-f%%. (a) Native TrE; (b) Trf control; (c) Tif-Adr conjugate. - OD?,,; ---- OD,9,.
mgiml) at 4°C for 30 min, followed by incubation with Trf-Adr, Cer-Adr, or PBS at 4°C for 30 min, they were then washed twice at 4°C and further incubated at 37°C for 2 hr. For the quantitation of cellular proliferation, 300 ~1 of complete medium was added to the washed cell pellet, and the cells were resuspended and distributed in triplicate as 100~l.~l aliquots in flat-bottom microtiter plates (Flow Laboratories), each well of which was preloaded with 100 ~1 of complete medium and 25 ~1 of 2 @Yi of [3H]thymidine (sp act., 2 Ci/mmol, Radiochemical Center, Amersham, Bucks., England). Plates were then incubated at 37°C in 5% C&/air for 16 hr, and the cells were harvested on glass fiber filters in a Mash II multiple harvester (Microbiological Associates, Walkersville, Md.). The amount of i3H]thymidine incorporated into DNA was measured in a LKB 1216 Rackbeta II liquid scintillation counter and expressed as counts per minute (cpm). Quantitative Ifleasures of the viability and number of cells were also done in all experiments.RESULTS Chromatographic Properties of Trfi Trf-Adr
Conjugates, and Adr
In preparing the conjugates, we found that native Trf (1 ml of 5 mglml in P consistently emerged from a column of Sepharose CL-6B between fractions 22 and 30, peaking at fractions 26 (Fig. la). However, when Trf, which had been carried through the conjugation procedure in the absence of Adr, was chromatographed on the same column, a more complex pattern was obtained (Fig. lb). The macromolecular peak (A in Fig. lb) was precipitated by anti-Trf antibody in Buchterlony and immunoelectrophoresis gels, as did the samples which emerged
4
YEH
AND
FAULK
FIG. 2. Immunological identification of Tti on cells and fluorescence localization of Adr in nuclei. Note uniformly speckled pattern of membrane (green) fluorescence produced by FITC anti-Trf (a), nuclear (red) fluorescence produced by Adr (b), and cells with both membrane and nuclear fluorescence following incubation with Tif-Adr conjugates (c). (X 625).
at the same place as native Trf, i.e., between fraction numbers 20-26 (B in Fig. lb), but the late appearing peak (C in Fig. lb) was not precipitated by anti-Tlf in gels. For the purposes of this study, “A” was considered to be Trf aggregates, “B” was chromatographically and antigenically native Trf, and “C” was Trf fragments. When Trf- Adr conjugates were passed through the column, the OD,,, tracing was similar to the controls, but reading at ODdg, revealed Adr in both peaks A and B (Fig. lc), whereas nothing was detected at OD,,, in the absence of Adr. Inasmuch as peak B was antigenically native Tti and its spectrophotometric characteristic indicated the presence of Adr, it was used as Tif-Adr conjugates in the following experiments. It should be noted that Adr has a molecular weight of 579.98 (23), and that it does not chromatograph within the molecular weight range of Sepharose CL-6B.
KILLING
TUMORS
WITH
FIG.
CYTOTOXIC
TRANSFERRIN
2-Continued.
Immunological IdentiYcation of Trf on Plasma Membranes and Fluorescence Localization of Adr in Nuclei
Trf, Adr, and Trf-Adr conjugates were investigated for their ability to be bound by plasma membrane receptors and to intercalate with nuclear DNA (24). This was done by using two different manipulations: the first being the incubation of Trf receptor-positive cells with the above reagents for 20 min at 4°C and the second being incubation of cells with the same reagents for 3 hr at 37°C. For these experiments, Daudi and HL-60 cells were used, since we have previously reported that their plasma membranes have receptors for Tif (7, I@, and PB were used as Tif receptor-negative cells (3, 4, 7-9). When HL-60 or Daudi cells were incubated with Trf which had been carried through the conjugation and chromatography in the absence of Adr, or with Ti-f-Adr conjugates for 20 min at
6
YEH
AND
FAULK
4°C a uniformly speckled pattern of fluorescence could be identified on their plasma membranes following incubation with FITC-labeled antibodies to human Trf (Fig. 2a). In contrast, only 8% of PBM were found to bind Trf or Trf-Adr conjugates, and none of the PBM, HL-60, or Daudi cells reacted with FITC antiTrf after being incubated with Adr alone (Table 1). However, incubation of PBM, HL-60, or Daudi cells with free Adr for 20 min at 4°C resulted in a red fluorescence of their nuclei (Fig. 2b), a result which was not obtained by incubation of any of the target cells with Trf or Trf-Adr conjugates for 20 min at 4°C. When these experiments were repeated with the use of the same cells and the same reagents, but with the incubation time changed to 3 hr and the temperature changed to 37°C the results were different. In the first instance, the pattern of immunofluorescence following incubation of HL-60 or Daudi cells with Trf or Trf-Adr conjugates was no longer homogeneous, but appeared to be clustered into islands of fluorescent patches on plasma membranes (Fig. 2~). The most striking difference found with these changed conditions of time and temperature for incubation was that with the Tif-Adr conjugates, for Adr could now be identified in the nucleus (Fig. 2~).
Effects of Trf-Adr
Conjugates
on [‘Hjthymidine
Uptake and Viability
of Cells
It is thought that Adr acts by inserting itself between base pairs of DNA (25), resulting in an inhibition of DNA synthesis and eventual cell death, hence we chose the inhibition of [3H]thymidine uptake as well as trypan blue exclusion as measures of cellular proliferation and viability, respectively. Incubation of TrfAdr conjugates for 3 hr at 37°C with Daudi and HL-60 cells followed by two washes and a 16-hr culture period in tritiated thymidine had the striking effect of diminishing both proliferation and viability, whereas minimal effects were observed for both thymidine incorporation and viability when PBM were tested under the same conditions (Table 2). The Trf control manifested moderate suppression of Daudi and HL-60 cells and augmentation of PBM proliferation, while Tif-Adr conjugates consistently recorded more than 90% suppression of both cell lines in 30 experiments. Regarding the death of PBM, it should be pointed out that, unlike Daudi and HL-60, very few PBM were found to contain Adr in their nuclei, indicating that dead cells resulted as a consequence of manipulation in vitro. In order to provide additional evidence for the specific inhibition of cellular proliferation by Tif-Adr conjugates, Cer-Adr conjugates were used as a protein control for Trf. The choice of Cer was predicated by the observation that other proteins such as human albumin, Gc, and alpha fetal protein are immunologically cross-reactive with unfolded Tif but not with Cer (26, 27). Although Cer-Adr conjugates were found to kill target cells in vitro, when Trf-Adr or Cer-Adr conjugates were allowed to displace Trf, Trf-Adr conjugates were almost twice as efficient as Cer-Adr conjugates (Table 3). Native Ti-f alone in the growth media
KILLING
TUMORS
WITH CYTOTOXIC TABLE
PLASMA
MEMBRANE
Reagents
TRANSFERRIN
7
1
AND NUCLEAR REACTIVITY OF Trf, Adr, AND Trf-Adr CONJUGATES CELLS AT DIFFERENT TEMPERATURES AND TIMES OF INCUBATION
Temperature (“C)
Tlf A& Trf-Adr Trf Adr Trf- Adr
Membrane fluorescence for Tif
Time
4 4 4 37 37 37
20 min 20 min 20 min 3 hr 3 hr 3 hr
a Percentage of cells fluorescing/intensity b Percentage of positive cells.
70-SO/3 0 70-8013 70-8013 0 70-SO/3
WITH
HL-60
Nuclear fluorescence for Adr
+n
0 1OOb 0 0 100 70-80
+ + +
of fluorescence (Ref. 18).
could not have been solely responsible for these results, since test and control experiments contained the same amount of Trf (Table 3). Dose and Time Variables
in Cell Killing
by Trf-Adr
Conjugates
Ti-f-Adr conjugates were serially diluted to test the endpoint of their ability to inhibit the uptake of [3H]thymidine. The results of these experiments showed that the proliferation of HL-60 cells was directly associated with the amount of TrfAdr conjugates used in the initial 3-hr preincubation step of the assay (Fig. 3). The percentage inhibition of cellular replication was rapid and achieved a plateau at 30 min preincubation (Fig. 4), although the in vitro assay for inhibition of cellular proliferation employed a preincubation step that was studied by using a xed amount of conjugates with varied periods of time.
TABLE 2 EFFECTS OF
Trf-Adr
ON [3H]T~~~~~~~~
UPTAKE
AND VIABILITY
[3H]thymidine uptake (cpm) Cellsb Daudi HE-60 PBM
OF CELLS
Viability (W
Control
Trf control
Trf- Adr
Control
Trf control
Trf- Adr
155,511 +- 6,793” 199,918 r 7,489 316 r 87
72,339 2 3,159 32,058 k 1,899 672 IL 215
14,850 i 812 3,515 i 124 98 r 25
89 92 94
73 88 87
63 56 86
a Mean 2 SEM from data of triplicate cultures obtained from 30 experiments. b Cells (0.5 x 106) were incubated with growth medium (control), Trf control (B in Fig. lb), or TrfAdr (B in Fig. lc) for 3 hr at 37”C, washed, and incubated with [3H]thymidine for 16 hr at 37”C, after which the uptake of [3H]thymidine and viability were determined.
8
YEH
AND
TABLE COMPETITIVE
INHIBITION
OF
Trf-Adr UPTAKE
FAULM
3
AND Cer-Adr WITH Trf ON [3H]T~~~~~~~~ OF DAUDI CELLS
Condition Reagent Tif- Adr Cer-Adr Control
No Preincubation
Preincubation
87,000 i 5,857 80,687 f 2,025 145,987 f 6,266
with Trf
48,624 + 3,724 85,563 i 5,677 122,467 ? 7,625
Note. A composite of results from two experiments in which cells were preincubated with or without 100 ~1 of Trf for 30 min at 4”C, followed by incubation with Tif-Adr. Cer-Adr, or PBS for 30 min at 4”C, washed twice at 4”C, and then further incubated for 2 hr at 37°C. The measurement of incorporation of [3H]thymidine was as detailed under Materials and Methods.
DISCUSSION
It has been proposed that some cancer chemotherapeutic agents might be surreptitiously transported into cells as conjugates of Tif through the mediation of Trf receptors (5), and the findings in this report support this idea by showing a temperature-dependent internalization of Ti-f-Adr complexes into the cytoplasm and the subsequent movement of Adr into the nuclei of Daudi and HL-60 cells. One explanation of these results was that TI-f-Adr conjugates were bound by Trf receptors on the cells, and following the 3-hr incubation at 37”C, they were endocytosed into so-called receptosomes (28). In consequence, the covalent bond between Tif and Adr in the conjugates was cleaved in the acidic cytoplasmic compartment (17) to liberate free Adr to diffuse into and be bound by nuclear DNA. The observed cytotoxic effect was directly associated with the amount of conjugate used, as shown by dose-response experiments, and this effect achieved a plateau following an incubation period of 30 min with the conjugate.
FIG. 3. Dose-response HL-60 cells.
of inhibition
of Trf-Adr
conjugates on the uptake of i3H]thymidine
by
KILLING
TUMORS
WITH CYTOTOXIC
9
TRANSFERRIN
20 i’ ’ 1;
t----
,
10
30
FIG. 4. Time course of inhibition cells.
60
of Trf-Adr
,
90 Pre-incubation
120
1
150
180
conjugate on the uptake of [3H]thymidine on Daudi
It has also been proposed that retinoic acid, which induces the differentiation of HL-60 cells, offers another approach to the treatment of some human leukemias (29). Indeed, following induction by retinoic acid or dimethyl sulfoxide, we have shown that the morphological and functional maturation of HL-60, but not of Daudi cells, was accompanied by a loss of Tif receptors (18), thus alterations in both cell types following exposure to Trf-Adr conjugates in the present experiments could not be due to a mechanism similar to that observed with either retinoic acid or dimethyl sulfoxide. Interestingly, ricin conjugates of monoclonal antibodies to Tif receptors or the antibodies alone have also been reported to inhibit the growth of human tumor cells in nude mice (30), again indicating that Trf receptors are present and are useful in studies of the biology of Trf receptors. Conjugates of Tif with Adr would not be expected to generate clinical problems, for Trf is a normal constituent of plasma, and the literature contains many reports of Adr chemotherapy without immunopathological sequelae, suggesting that the use of Trf-Adr conjugates would probably not cause untoward immunological reactions in patients. Furthermore, when leukemic cells were incubated with TsfAdr conjugates in the presence of normal human plasma, the percentage of inhibition remained unchanged (unpublished data), suggesting that the Trfreceptors are available to bind cytotoxic Ti-f from the blood circulation. It is possible that the Trf receptor is a family of structurally related but antigenitally and functionally distinct molecules much as was suspected as a result of studies of breast adenocarcinoma cells (5). This is suggested by observations that some but not all monoclonal antibodies to Trf receptors recognize receptor in human skin (31, 32). Thus, physiological receptors for Trf on normal cells may be irrelevent in our proposed regimen. There are two additional usages of Trf and Trf receptors which could be exploited: the first being that radiolabeled Trf might be used for the detection of primary or metastatic tumors as is now done by using
10
YEH AND FAULK
Gallium (33) or isotopically labeled antibodies to carcinoembryonic antigen (34); and second, the possibility of preparing a battery of conjugates of Trf with other chemotherapeutic agents to assess the relative efficiency of each to kill bone marrow cells in vitro from newly diagnosed leukemic patients as an adjunct in deciding what agent should be employed for therapy. The latter possibility of drug tailoring to fit the patient is supported in part by unpublished data from our laboratory showing that PBM and bone marrow cells from patients with acute myelogenous leukemia were selectively killed in vitro by Trf-Adr conjugates in a fashion similar to that described in this report. The results of these experiments form the basis for a subsequent report. ACKNOWLEDGMENTS This research was supported in part by the British Medical Research Council, INSERM Unit 210, the East Grinstead Research Trust, and the Human Embryology and Pregnancy Foundation. We thank Dr. M. Steinbuch for her gift of human ceruloplasmin.
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
Fauik, W. P., and Galbraith, G. M. P., Proc. Roy. Sot. Lorzdon B 204, 83, 1979. Wada, H. D., Hass, P. E., and Sussman, H. H., J. Biol. Chem. 2.54, 12629, 1979. Larrick, .I. W., and Cresswell, P., Biochim. Biophys. Acta 583, 483, 1979. Galbraith, G. M. P., Galbraith, R. M., and Faulk, W. P, Cell. Immunol. 49, 215, 1980. Faulk, W. P., Hsi, B.-L., and Stevens, P. J., Lancet 2, 390, 1980. Shindelman, J. E., Qrtmyer, A. E., and Sussman, H. H., Int. J. Cancer 27, 329, 1981. Yeh, C. G., Hsi, B.-L., and Faulk, W. I?, In “Aspects of Developmental and Comparative Immunology” (J. B. Solomon, Ed.), 361, Pergamon, Oxford, 1980. Larrick, J. W., and Logue, G., Lancet 2, 862, 1980. Yeh, C. G., Taylor, C. G., and Faulk, W. P., VOX Sang&us, in press. Gmary, M. B., and Trowbridge, I. S., J. Biol. Chem. 256, 12888, 1981. Schneider, C., Sutherland, R., Newman, R. A., and Greaves, M. F., J. Biol. Chem. 2.57, 8516, 1982. Enns, C. A., Suomalainen, H. A., Gebhardt, J. E., Schroder, J., and Sussman, H. H., Proc. Natl. Acad. Sci. USA 79, 3241, 1982. Goodfellow, l? N., Banting, G., Sutherland, D. R., Greaves, M. F., Solomon, E., and Povey, S., Somat. Cell Gent. 8, 197, 1982. Sullivan, A., Grasso, J. A., and Weintraub, L. R., Blood 47, 133, 1976. Hemmaplardh, D., and Morgan, E. H., Brit. J. Haematol. 36, 85, 1977. Karin, M., and Mintz, B., J. Biof. Chem. 256, 3245, 1981. van Renswoude, J., Bridges, K. R., Harford, J. B., and Klausner, R. D., Proc. Natl. Acad. Sci. USA 79, 6186, 1982. Yeh, C. G., Papamichail, M., and Faulk, W. P., Exp. Cell Res. 138, 429, 1982. Pegaudier, Z., Andran, R., and Steinbuch, M., Clin. Chem. Acta 30, 387, 1970. Van Vunakis, H., Langone, J. H., Riceberg, L. J., and Levine, L., Cancer Res. 34, 2546, 1974. Hurwitz, E., Levy, R., Maron, R., Wilchek, M., Amon, R., and Sela, M. Cancer Res. 35, 1175, 1975. Yeh, C. G., Hsi, B.-L., and Faulk, W. P., J. Immunol. Methods 43, 269, 1981. Arcamone, E, Cassinelli, G., Francheschi, G., Pence, S., Pol, C., Redaelli, S., and Selva, A., In “international Symposium on Adriamycin” p. 9, Springer-Verlag, Berlin/Heidelberg/New York, 1972. Pigram, W. J., Fuller, W., and Hamilton, L. D., Nature (London) 217, 235, 1972. Zunino, E, Gambetta, R., DiMarco, A., and Zaccara, A., Biochim. Biophys. Acta 277,489, 1972. Pekkala-Flagan, A., and Ruosiahti, E., J. Immunol. 128, 1163, 1982.
KILLING
TUMORS
WITH CYTOTOXIC
TRANSFERRIN
If
27. Metz-Boutique, M. H., Jolles, J., Mazurier, J., Spik, G.; Mantreuil, J., and Jolles, P., FEBS Letf. 132, 239, 1981. 28. Willingham, M. C., and Pastan, I., Cell 21, 67, 1980. 29. Breitman, T. R., Selonick, S. Z., and Collins, S. J., Proc. Natl. Acad. Sci. USA 77, 2936, 1980. 30. Trowbridge, I. S., and Domingo, D. L., Nature (London) 294, 171, 1981. 31. Wells, M., Yeh, C. G., Hsi, B.-L., and Faulk, W. P., J. Clin. Pathol., in press. 32. Gatter, K. C., Brown, G., Trowbridge, I. S., Woolston, R. E., and Mason, 0. Y., J. Clin. Pathoi. 36, 539, 1983. 33. Silberstein, E. B, C/in. Nucl. Med. 6, 63, 1981. 34. Mach, J. P., Oncodev. Biol. Med. 1, 49, 1980. Received July 6, 1983; accepted with revision January 12, 1984.