Cotreatment or pretreatment of several human myeloid cell lines (KGI, HL60, U937, THP1) with the diferentiation inducer DMSO was found to potentiate the antiproliferative and cytotoxic effects of TNF. In addition, TNF-resistant monocytic cell lines could be sensitized to TNF cytotoxicity by DMSO treatment. Other highly polar molecules, known to be potent differentiation inducers, showed similar effects to those of DMSO. The potentiating effect of DMSO was related neither to an up-regulation of TNF receptor expression nor to an alteration in the rate of TNF internalization and degradation. We present evidence that the TNF activities are p55 TNF receptor-mediated and are not due to insertion of TNF into lipid bilayers, an effect that could be susceptible to DMSO, as this component has been described to modify cell membrane characteristics. DMSO-induced potentiation of TNF cytostasis/cytotoxicity was restricted to myeloid leukemia cell lines. In non-myeloid cells such as fibrosarcomas, myosarcomas, thymomas, or carcinomas, DMSO was found either not to alter or to inhibit TNF-induced cell death. The latter results are in good agreement with data reported by others who suggested that DMSO could act as a scavenger of TNF-induced toxic radical formation. The potential correlation in mycloid cells between DMSO-induced changes in the cells' differentiation status and DMSO-enhanced TNF-susceptibility is discussed.

J. Leukoc. Biol. 57: 141-151; 1995.

DMSO	dimethyl sulfide
IFN-	interferon
IL	interleukin
mAb	monoclonal antibody
TNF	tumor necrosis factor
U937r	TNF-resistant U937 cells

Tumor necrosis factor (TNF) was originally described on the basis of its cytostatic and cytotoxic effects on tumor cell lines, and was found to be present in the serum of animals injected with endotoxin . ¹ The cDNAs for human and mouse TNF have been cloned and expressed in prokaryotic or eukaryotic systems ,^{2, 3} resulting in the availability of highly purified recombinant TNF in sufficient quantities for detailed in vitro and in vivo studies. The protein is recognized to be a cytokine with pleiotropic biological properties. Besides its cytostatic and cytotoxic effects on certain tumor cell lines ,⁴ TNF influences growth, differentiation, and/or function of virtually every cell type investigated (for review, see refs. 5 and 6). The action of TNF requires specific binding to high affinity cell surface receptors of which two types are known, namely p55 and p75, named according to their respective molecular weight . ^7, 8 Most cell types express both types of TNF receptors, albeit to a different relative extent . ^9, 10

Several agents have been described to potentiate TNF activity. For example, susceptibility to TNF cytotoxicity can be enhanced or induced by inhibitors of transcription or translation such as actinomycin D or cycloheximide respectively . ¹¹ A plausible explanation of this phenomenon is the inhibition of the expression of cellular rescue mechanisms. TNF effects can also be potentiated by other cytokines such as interferon  (IFN-),^12, 13 interleukin (IL)-l,^14, 15 IL-4,¹⁶ and IL-6. ¹⁷ Part of the potentiation of cell growth inhibition, observed in the case of IFN- and IL-6, could be due to the observed up-regulation of TNF receptor expression. ^12, 17 In contrast, IL-1 and IL-4 do not modulate^14, 16 or even down-regulate¹⁵ TNF receptor expression. Thus, besides TNF receptor modulation, other presently unknown intracellular pathways must be activated by these cytokines in order to exert their potentiating effects. Other up-regulators of TNF cytotoxicity are inhibitors of DNA topoisomerase II,¹⁸ heat treatment ¹⁹ inducers of intracellular cAMP accumulation,²⁰ lithium chloride,²¹ the protein kinase inhibitor staurosporine,²² and expression of the adenovirus E1 gene. ²³

Induce

We have studied human myelold cell lines for susceptibility to TNF-mediated cytostasis/cytolysis under various conditions. Some of these myelold cell lines provide excellent model systems to study control mechanisms of mammalian cell differentiation and function. Dimethyl sulfoxide (DMSO) and other polar agents induce these cells to differentiate into mature granulocytes or macrophage-like cells depending on the agent and the cell line used. ^24-27 Here we report that treatment with DMSO strongly up-regulates the susceptibility of several myelold cells to TNF-induced cell lysis. This potentiating effect was found to be specific for myeloid cells which mature towards a macrophage- or granulocyte-like phenotype upon incubation with DMSO.

TNF-induced effects are thought to be TNF receptor- mediated. ^28, 29 Recent reports, however, describe the ability of TNF to insert into lipid bilayers and to exert biological activities via a non-receptor-mediated pathway. ^30, 31 As DMSO is believed to exert changes in cell membrane characteristics. ^32, 33 TNF insertion could also be implicated in the observed TNF effects in DMSO-treated cells. In contrast, we provide evidence that the TNF cytoxic effects are TNF receptor-mediated, although DMSO did not profoundly alter the TNF receptor characteristics.

Reagents

Recombinant human TNF (specific activity: 5.10⁸ U/mg protein in the WEHI 164 cl13 cytotoxic assay³⁴ was prepared as described. ² Mutants of human TNF showing reduced binding to human p55 TNF receptor were prepared as described by Van Ostade et al. ³⁵

Monoclonal

All polar compounds used were from Aldrich (Milwaukee, WI) except for DMSO which was from Merck (Darmstadt, Germany). Mouse monoclonal antibodies (mAbs) directed against the human p55 or p75 TNF receptor¹⁰ were generously provided by Dr. M. Brockhaus (Hoffinann-La Roche, Basel, Switzerland).

Cell lines and culture conditions

The following human myeloid leukemia cell lines were used: the erythroleukemia cell line K562,³⁶ the very early myeloblast KG1,³⁷ the promyelocyte HL60,³⁸ the myelomonocytes U937³⁹ and THP1,⁴⁰ and the monocyte MonoMac6. ⁴¹ The fibroblastic NIH3T3 cells were originally from Dr. R. Weinberg (Whitehead Institute for Biomedical Research, Cambridge, MA). The 24T2·5 cell line was from Dr. H.  Schreiber (University of Chicago, Chicago, IL). The murine L929 and WEHI 164 cI13 fibrosarcoma cells were from Dr. R. Konings (Rega Institute, Leuven, Belgium) and from Dr. T. Espevik (University, of Trondheim, Trondheim, Norway³⁴), respectively. KYM39A6 is a subclone of the KYM-1 rhabdomyosarcoma cell line obtained from Dr. A. Meager (National Institute for Biological Standards and Control, South Mimms, UK⁴²). Transfection of human TNF receptor p55 and p75 CDNA in the rat/mouse T cell hybridoma PC60 (PC60 TR55/75) was performed as described. ⁴³ All other cell lines used were from the American Type Culture Collection (Rockville, MD).

All cell lines were maintained in RPMI 1640 medium supplemented with 10% (v/v) heat-inactivated calf serum, 2 mM L-glutamine, 100 U/mi penicillin and 100 mg/ml streptomycin. Culture medium for the MonoMac6 cell line was further enriched with 1 mM pyruvate, nonessential amino acids, 1 mM oxaloacetate and 9 g/ml insulin. All tissue culture reagents were from Gibco (Paisley, Scotland).

A TNF-resistant variant of the U937 cell line (U937r) was selected by culture of the parental U937 in the presence of 10 ng/ml TNF for several months until they grew stably under selective pressure (B.V., unpublished data). Before use, the U937r were cultured in the absence of TNF for at least 3 days.

For DMSO pretreatment, exponentially growing cells were harvested by centrifugation (10 min at 800g), resuspended at 2 x 10⁵ cells per ml fresh culture medium supplemented with the indicated concentrations of DMSO, followed by an incubation at 37°C for different time periods. For further use, cells were collected by centrifugation and DMSO was removed by three washes in RPMI 1640 medium. In case of adherent cell populations, cells were detached from the plastic surface by means of a policeman.

Assays of cell growth and cell viability

Cell proliferation was determined by [³H]thymidine incorporation into DNA. Ten thousand cells were cultured in flat-bottomed 96-well microliter plates (Nunclon, Roskild, Denmark) in presence of the indicated agent for 24 h. [³H]thymidine (25 Ci/mmol; Amersham International, Amersham, UK) was added at 0.5 Ci/well for the last 4 h of the 24-h incubation period. Cells were harvested with an automated cell harvester and incorporated radioactivity measured by liquid scintillation counting. Each experiment was performed in triplicate.

Because a decrease in thymidine incorporation may simply reflect cytostasis and not cytotoxicity, parallel experiments were set up each time by seeding cells under identical conditions in 24-well plates. After 24 h, the number of viable cells were assayed by trypan blue staining. In some experiments, TNF-induced cytostasis/cytotoxicity was determined by the calorimetric MTT assay. ⁴⁴ Briefly, 2 x 10⁴ cells were cultured in 96-well microliter plates in presence of TNF with (in case of the adenocarcinoma HFp2 cell line) or without 50 g/ml cycloheximide. After 72 h (24 h in case of assays with cycloheximide), cells were stained with MTT.

Receptor binding experiments

Cells (2-3 x 10⁶) were incubated with radioiodinated human TNF (600-800 Ci/mmol; Amersham) in a total volume of 300 l in Eppendorf tubes. RPMI 1640 medium supplemented with 10% fetal calf serum was used as the binding buffer. After 2 h incubation on ice, cells were centrifuged and washed three times with ice-cold binding buffer and the radioactivity of the cell pellet was measured in a gamma counter (Packard Instrument Company, Meriden, CT). Non-specific binding, which was determined in the presence of a 100-fold excess of unlabeled TNF, never exceedcd 10% of total binding (specific binding = total binding - non-specific binding). All binding assays were performed in duplicate and the data were subjected to Scatchard analysis.

In receptor blocking experiments, cells were incubated for 1 h at 37°C with 20 g anti-TNF receptor antibodies per ml. Unbound antibodies were washed away before additon of [¹²⁵I]THF.

Internalization and degradation of cell-bound [¹²⁵I]TNF

Cells were allowed to bind saturating concentrations of [¹²⁵I]TNF at 4°C as described above, after which total cell-bound radioactivity was measured. Thereafter, cells were incubated with fresh prewarmed (37°C) binding buffer. Kinetics of internalization and degradation were further assayed at 37°C. After different incubation times, cells were separated from the supernatant by centrifugation. The cell pellet was used to determine cell surface-bound and internalized [¹²⁵I]TNF, whereas the supernatant was used to monitor dissociation and degradation of [¹²⁵I]TNF. For determination of cell surface-bound and internalized TNF, cells were rinced in RPMI 1640 medium and incubated for 5 min at 4°C with 1 ml 0.05 glycine HCl pH 3.0, 0.15 M NaCl, after which the acid-removable (cell surface) and the acid-non-removable (internalized) fraction of [¹²⁵I]TNF were counted. For the determination of dissociated and degraded TNF, cell dissociated TNF in the supernatant was treated with 10% trichloracetic acid for 5 min at 4°C and then centrifuged at 1000g for 30 min. Both acid-soluble (degraded) and -insoluble (non-degraded, cell dissociated) fractions were counted.

DNA isolation and electrophoresis

The number of cells seeded at the beginning of the treatment differed depending on the growth kinetics and/or TNF sensitivity of the cell line under investigation. However, at the time of harvest and subsequent DNA isolation, approximately 10⁶ cells (dead and/or alive) were present in each culture. Cells were washed twice with phosphate-buffered saline by a 5-min centrifugation at 8000 rpm in an Eppendorf microfuge. Cell pellets were resuspended in cell lysis buffer (0.1M NaCl, 40 mM Tris HCI pH 7.2, 20 mM EDTA, 0.5% sodium dodecyl sulfate, and 200 mg/ml proteinase K) and rotated overnight at 37°C. The crude DNA preparations were extracted twice with phenol, followed by a chloroform/isoamylalcohol (24:1, v/v) extraction. After the addition of 5 M NaCl (1/10 of volume) and 2 volumes 100% ethanol, the DNA fraction was precipitated overnight at -20°C. After centrifugation at 14,000 rpm for 15 min, the DNA pellet was dissolved in 25 l Tris-EDTA buffer (10 mM Tris HCL pH 7.6, 1 mM EDTA) and RNAse A was added to a final concentration of 100 mg/ml. After 2 h incubation at 37°C, the DNA samples were loaded on a 1.8% agarose gel in the presence of bromophenol blue. Electrophoresis was performed at 90 V in TAE buffer (40 mM Tris acetate pH 7.9, 2 mM EDTA), after which DNA was visualized by ethidium bromide staining.