TNF Receptors TR60 and TR80 Can Mediate Apoptosis Via
Induction of Distinct Signal Pathways 1
Matthias Grell,'* Gudrun Zimmermann,. Dieter HUlser,+ Klaus pfizenmaier,. and
Peter Scheurich*
-Institute of Cell Biology and Immunology, University of Stungart, Stuttgart, Germany; and tBiological Institute, Department
of Biophysics, University of Stuttgart, Stuttgart, Germany
TNF membrane receptors are usua lly co-expressed in many tissues but their relative contribution to cell ular TNF
responses is for most situations unknown . In a TNF cytotoxicity model of KYM·l, a human rhabdomyosarcoma cell
line, we recently demonstrated that each of the two TNFRs is on its own capable of inducing cell death. Here we
show that both receptors are able to induce apoptosis, as revealed (rom a similar onset of DNA fragmentation and
typical morphologic criteria. To obtain additiona l information about the signa ling pathways involved in TRoO· and
TR80· induced programmed cell death, we have used a series of selective inhibitors of intracellular signaling
molecules. The overall pattern emerging from these experiments provides strong evidence for distinct signal
pathway usage of TR60 and TR80, indicating protein kinase(s)·mediated control of TR60 signaling and a tight
linkage of TR8Q to arachidonate metabolism . The subsequent establishment of KYM·1·derived cell lines that
display TNFR selective resistance further supports a segregation of TR60 and TR8Q signaling pathways for induc·
tion of apoptotic cell death. Moreover, these results demonstrate an independent control of the distinct signaling
cascades used by TR60 and TR80. This allows a highly flexible regulation of a cellular TNF response in those cases
in which both receptors contribute to overall TNF responsiveness. The Journal of Immunology, 1994, 153 : 1963,
T
NF is a pleiotrophic cytokine that is primarily pro-duced by activated macrophages. Originally
named for its ability to induce tumor necrosis in
certain model systems, TNF is now known to play a major
role in many processes of inflammation affecting both he-matopoietic and nonhemopoietic tissues. Beside a physi·
ologic role of TNF in ontogeny and control of certain
infections, TNF is also recognized as an important patho·
genic factor in several chronic diseases (for reviews see
Refs. 1, 2). This dual role of TNF in vivo is on the basis
of the ability of TNF to modulate the expression of a va·
riety of genes, including several other proinfiammatory cy·
tokines such as If......l , If......6, granulocyte·macrophage CSF,
cytokine receplors, adhesion mol ecules, and enzymes of
various metabolic and catabolic pathways (reviewed in
Received 'Of publiution Febnlary 4, 1994. "cceplt<! fOf publication May 31,
1994.
The cost!; of publ ication of this iutide were defrayed in part by the pavmcnl of
page chasgcs. This ~rtide musI the<tforc be hereby ~fked ~;~f in
accordance wilh 18 U.S.c. SeaiOn 1734 ~lely to indiCilte this filet.
, This work was supported by DetJlSChe F~hungsgemeinschaft. Crant Sche
349/1-).
J Address COfrespondcnce and ~nt requests to Dr. Manh~ Crell, Institute
of Cell Biology and Immunology, Universi(yof Stultprt. Allmandri na )1 .
70569 Stul\gilrt. Germany.
Copyright 0 1994 by The American Association of Immunologilols
Ref. 3). Furthermore, TNF has been demonstrated 10 pas·
sess direct cytotoxic activity in vitro for certain tumor cells
and for some nonnal cells (4-6).
TNF and Iymphotoxin initiate their broad range of eel·
lular responses by interaction with cell sudace membrane
receptors. Two distinct but related receptor molecules
have recently been molecularly cloned (7-10), here re·
ferred to as TROO (type I) and TR80 (type 11). Both re·
ceptors have significant homologies in their extracellular
domains with repeat cysteine-rich sequences, defining
them as members of a recently established growing recep·
tor family (11). In contrast, the cytoplasmic domains of
TR60 and TR80 are unrelated and give no indication by
whlch mechanisms they are coupled to and activate intracell ular signaling pathways,
In addition to differences in primary structure. expression of the two TNFRs seems to be differentially regulated
and shows lissue·spccific prevalence (1Z-14), Thus, although C<rexpression of both receptors is found in many
cells, there is typically a quantitative dominance of one of
the two receptors. For example, lymphoid cells predomi·
nantty express TR80 molecules, whereas epithelial cells
typically express TR60 (15).
Concerning the cytotoxic activity of TNF, recent findings have drawn new attention to this particular action as
OCI22· 1767!)4IS02.00
1964
INDUCTION OF APOPTOSIS VIA BOTH TNF RECEPTORS
an important physiologic function. First. a role of TNF as
a possible stimulating and effector molecule of NK/lymphokine-activatcd killer cells is apparent from several
studies (16, 11). Second, TNF seems to function as a tranSducer of cellular cytotoxicity in its membrane-integrated
form, c.g., on macrophages (IS). Third. and most intriguing,
recently, TNF has been proposed to act as an inducer of apoptotic ceU death during thymocyte maturation (6, 19).
Qearly. 1NF has the polentialto exert cytotoxic effects for
different cell types in differential ways, because it can induce
both necrotic and apoptolic forms of cell death (20, 21).
The cellular events in lNF-mediated cytotoxicity after
ligand binding to membrane receptors are still poorly understood. Involvement of different signal transduction
pathways has been suggested. These include activation of
phospholipase (PL») A2 with release of arachidonate melaboliles (22. 23). Slimulalion of PLs D (24) and C (25.26)
with activation of protein kinase C (PKC) by diacylglycerols, activation of sphingomyelinase activity with subsequent release of ceramides (27), and, finally, hydroxyl radical production by mitochondrial enzymes (28). On the
other hand, most tumor cell lines are sensitive to the cytotoxic action of TNF only in the presence of synergizing
biologic or chemical reagents. This seems to be true independent of the pattern of TNFR expression. Therefore, it is
likely that intracellular regulatory circuits control the sensitivity toward the cytotoxic action of TNF. For example,
IL-l and TNF, itself, have been demonstrated to induce
TNF resistance in a protein synthesis-dependent manner
(29). Key molecules in induction of lNF resistance could
represent manganese-dependent superoxide dismutase
(30). heal shock proleins (31). and PKs (32. 33).
Until recently the role of the two TNFRs for induction
of cytotoxic effects of TNF was unclear. As of now, the
vasl majority of cellular responses have been attributed 10
signaIing via TR60, whereas only a few examples exist to
demonstrate TRSO-initiated cellular answers. In particular,
TNF-mediated cytotoxicity has been linked to TR60 in a
number of different buman cell lines (12, 34). The principal capability of TR60 to trigger cytotoxicity on its own
has been clearly demonstrated in heterologous human!
mouse bioassays in which a selective cross-species acti·
valion of the murine TR60, but not TR80, was achieved
(35). Furthermore, studies with receptor subtypc-specific
agonistic antisera have indicated that Abs specific for murine TR60 induce cytotoxicity. whereas otherwise agonistic TR80-specific sera failed to do so (36).
A recent report has attributed this capability of TR60 to
a cytoplasmic region with homology to the apoptosis inducing Ag FaslAPO·l, another member of the TNFR family (31). Although these data suggest that TR60 plays a
major role for induction of a cytotoxic response, other
studies indicate that TRaO may also be involved. Thus, in
U937 cells, tbe blocking of each of the receptor subsets
resulted in a significant reduction in lNF-induced cytotoxicity (38). Furthermore, it was shown that overexpression of TR80 in HeLa cells strongly enhanced TNF sensitivity (39). An even more complicated situation might be
found in thymocytes in which the pattern of cytotoxic responses to TNF could be controlled by both receptors in a
very subtle way (6).
Using the human rhabdomyosarcoma cell line KYM-l,
which co-expresses high numbers of both receptors, we
recently demonstrated that each receptor is able to induce
cytotoxicity on its own (40). Moreover. limited receptor
triggering that used the respective specific Abs indicated
additive action (40). Here we show that, on cross-linking,
each receptor induces apoptosis, as revealed from the induction of DNA fragmentation and typical morphologic
changes. To obtain additional infonnation about the signal
transduction pathways involved, we used a number of selective inhibitors to interfere with either TR6()... or TRSOinduced apoptosis. Furthermore, we established JCYMI-derived cell clones showing receplor-type selective
resistance. Both experimental approaches indicate clearly
that the two TNFRs initially activate different signal pathways that eventually merge and lead 10 the identical cellular response, i.e., apoptosis.
, Abbreviations u!ted in this p.;Iper: Pi. phospholipase; pc, phosphatidylcholine; Pie, protein kinase; ROI, radiul oxygen intennediate; BPB, brOfTlOphenKyl bromide; NCOC. 2·nilll>4-arboxyphl!nol N,N-<liphenylurbiilrmt~;
NDGA, nordihydroguairefic. ac.id; TIFA, thenoyhrinUOf03c.etone; PleC. protein
kina.5e C; PKA, protein kmase A.
in 1% osmium tetrolfide (Muck, Dannstadt. Germany) in 0.1 M PBS was
followed by dehy~tion in ethano t (8 min in 40, SO, 60, 70, and 8()11,
ethanol. 2 X IS min in 96 and 100% ethanol), and propytene oxide (2 X
IS min). After impregnation overnight with 2;1 and 1:1 mixtures of propylene oxide and the epoxy rc.sin, respectively, cells wc.re embedded in
Materials and Methods
Cell lines and reagents
The human rhabdomyosarcoma ceU line KYM-l (41) was generously
supplied by Dr. M. Sekiguchi (tnstil\llC of Medical Sc:ic.na:, University of
Tokyo, Tokyo, Japan). Cells wc.rccultured at3rC and 5% CO 2 in Oitles
RPMll64(1 medium (Biochrom. Berlin, Germany) conwning 10% beatinactivated FCS and antibiotics. The gtDCf81ion and specificity of the
mAb H398, directed against TR60 (12), the rabbit anti·human TR80
serum MOO (40), and the agonistic mAb htt·l (IS) has been described. To
generatc rcsislallt KYM· I subc:loncs, cells were grown under culture coo·
ditions in the presence of roc.asing concentrations of hlr-! for KYM·
60res or MBO serum for KYM·8Qres cells. Human rTNF-a (sp. aa. 2 x
10' U/mg) was kindly provided by Knoll AG, Ludwigsbafen, Ckrmany).
The following reagents were purchased from Slgma Chemical Co. (Dei·
senhofen, Germany); amyta1 (amobarbitat); antimycin; bromophcnacyl
bromide (BPS): genistein; lithium chloride, 2-nitr0-4-carboxyphenyl
N.N-diphenylcarbamate (NCDC): neom yc in; nordihydroguaiatetic acid
(NDGA); pargytine; quinacrine; and tbenoyltrifluoroaoelOnc (TrFA).
H7, H8, H89, K·2S2a, K-252b, staurosporine, and calphostin C were
from Calbiochem, Bad Sodc.n. Gennany. D609 was from Kamiya Biomedical Company. Thousand Oab, CA. For use in cytotoxicity assays,
stock solutions of tbe inhibitors were prepared in medium, ethanol, or
DMSO as appropriate, such that the final concentration never exceeded
0.5%, which was conlIolled not to affect cell viability or TNF
cytotoxicity.
Electron microscopic histology
Cell Cllltures were filed ill 2.5% g1utanldehyde (Merc.k, Darmstadt, Germany) in 0.1 M PBS at pH 7.1 for that room temperature. Postfixation
The Journal of Immunology
gelatin capsules and polymcriz.cd for 12 h at 4{j·C and at leas! 48 b at
70·C. Specime ns were thin sectiooed witb a diamood knife on a Reichert
O~ U) ultramicrotome (Re~chert.Jung, NUOlocll, Germany), stained
with uranyl acetate and lead Cllrate, and examined with a Zeiss EM lOA
electron miCfOSC(Jpt al 60 kV.
Detection of DNA fragmentation
Cells.were inc.ubaled under cullure conditions with the indicated reagents
for dl[ercnt limes, then harvested, and DNA was isolated by standard
procedur~ DNA was. ~al yzcd b~ electrophoresis on a 1.2% agarose gel
and visualized by elhiCfium bromide staining.
Cytotoxicity assay
This assay was conducted essentially as desaibed previously (5). Briefly,
cells we re seeded into 96-well microtiter plates al I X lit ce1lslw-ell and
aUowed 10 grow overnight before lhe addition of the different subslances
to a final volume of 200 ~. For inhibitor studies, the respcclive concentrations of lhe different substances nOI exhibiting more than 10% cytoloxic elrect. per se, over !be assay period were inilially determined. lbe
5ubs/;ances were added 30 to 60 min before lNF or the raxptor.specific:
agonists. Experiments Ihal used calphostin C as a specific PKC inhibitor
were performed under iUumination (42). After 18 b of culture MIT.
dimethylthiazol bromide (10 JLI/well of a S mg/ml solution in PBS) was
added 10 all wells and metabolically active cclls were allowed 10 produce
the formazan product for 2 h at 3rC. Subsequently, 90 ~ of a 15% SDS
solution in 0.02 N Ha was added. 00s were read after an additional 4·b
incubation at room lemperature at S60 om.
Results
Both TNFRs can independently mediate cytolysis in the
human rhabdomyosarcoma cell line KYM-l (40). To approach the molecular mechanisms involved, we first investigated, by using eleclron microscopic analyses, Ihe ultrastructural changes preceding lNF-induced cell death in
KYM- l cells. After stimulation wilh either TNF or the
respective receptor-specific agonistic Abs, the typical onsel of apoptotic cell death was revealed in all cases (Fig.
l). Accordingly, treatmenl wilh either TNF, Ihe TR60-specific mAb H39S plus a secondary cross-linking reagent
(goat anti-mouse IgG), or the polyclonal TRBO-specific
rabbit serum MSO (40) initiated strong vacuolization, condensation of chromatin at the nuclear membrane, and cellular fragmentation. All of these morphologic criteria have
been associated with programmed cell death (43). Additional support for the induction of apoptosis, rather than
necrosis, comes from studies of DNA fragmentation
caused by induction of an endonuclease(s) that preferentially digests internucleosomal DNA (44). As shown in
Figure 2, incubalion of KYM- I cells with TRSO-specific
agonistic Abs resulted in a typical DNA degradation pattern with approximately ISO-bp steps similar 10 the DNA
ladder observed on TNF-treatment or TROO-specific stimulation . Thus, independent and selective triggering of
1NFRs in each case resulted in typical apoptotic cell
death.
Because TRSO, like TR60, is capable of conferring apoptotic signals to the cell nucleus. we next asked whether
both receptors use idenlical intracellular signaling pathways. To determine this, we used several inhibitors of
well-defined signal transduction pathways that have been
1965
implicated in TNF responses and asked whelher they
would interfere with lNF-induced cytolysis of KYM-l
cells. Initially, we tested a broad variety of kinase inhibitors for thcir ability to interfere with 1'NF-mediated cytotoxicity. Neither the potent tyrosine kinase inhibitor
genistein nor the selective PKC inhibitor calphostin C
showed Significant effects on TNF-induced cytotoxicity. In
contrast, the isoquinoline s ulfonamide H8 clearly antagonized with the cylotoxic effects of TNF (Fig. 3). Only partial inhibition, however, was noted at the concentrations
used. Higher concentrations of this PK inhibitor could nOI
be used because of toxic side effects of the compound
itself. Interesti ngly, the related compounds HS9 and H7
did not significantly interfere with TNF cytotoxicity. In
contrast, the highly potent serine/threonine and tyrosine
PK inhibitor staurosporine, similar to the compounds
K252a and K252b, strongly enhanced TNF-induced
cytotoxicity (Fig. 3).
Radical oxygen intermediates (ROIs) have been shown
to be involved in lNF-induced cytotoxicity in different
cellular systems (28, 45. 46). Therefore, we investigated
several inhibilors of the mitochondrial electron transport
system for interference with TNF-mediated cytotoxicity in
KYM-l cells. Clearly, amy tal, a complex I inhibitor, exerted protective effects, whereas lTFA, an inhibitor of
complex II formation, was ineffective (Fig. 3). Antimycin
A. a complex 1II inhibitor, which leads to the accumulation of ROIs in the mitochondria (46), enhanced TNF cytOloxicity. The laller data are in accordance with the proposed role of oxygen radical generation as a causal
mechanism of TNF cytotoxicity (28) in KYM-l cells.
Regarding the other compounds tested, the Jipoxygenase inhibitor, NDGA, the PL inhibitor, neomycin, and the
inhibitor of monoamine oxidase. pargyline, all significantly inhibited TNF-mediated cytotoxicity in KYM-l
cells. However, in all cases only partial protection was
observed at the highest possible (nontoxic) concentrations
used (Fig. 3). These res ults are in agreement with reports
demonstrating the involvement of Pl..Az and lipoxygenase
metabolites in cell killing by TNF (23. 24). However. a
number of additional PtAz inhibitors used, such as BPS,
NCDC, or quinacrine, and the PC-PLC-specmc inhibitor
D609 failed to protect KYM-l cells from lNF-mediated
cytotoxicity, although thcy have been shown to interfere
with TNF signaling in other cellular systems (22, 27).
Some compounds that inlcrfcre with TNF action have been
shown to be effective via modulation of TNFR expression
(47, 48). To verify that the substances used here are indeed
effective at a postreceptor, i.e., signal transduction level, binding competition studies with radiolabeled TNF and the respective receptor-specific Abs were performed. These experiments revealed that none of the 1NF response-inhibiting
substances, Le., H8, neomycin, NDGA, pargyline, and
amytal. significantly changed the expression of the two
TNFRs or interfered with ligand or Ab binding to membrane
expressed receplors (data not shown).
1966
INDUCTION OF APOPTO$IS VIA BOTH TNF RECE PTORS
FI GU RE 1. Apoptosis of KYM-l cells induced via both TNFRs. Transmission electron micrographs were taken after 8 h of
treatment of KYM-l cells with medium (A), 20 nglml TNF (B), I :200 dilution of TRBO-specific rabbit serum M80 (q, or 3 p.g/ml
TR60-specific mAb H398 plus 10 p.glml goat anti-mouse IgG (0). Arrowheads indicate nuclear membrane without (Al or with
condensed chromatin ( 8 through D). Ba rs " 5 }J.m .
Cu rrently, all of the known TNF signal Iransduction
pathways have been linked to TR60 rather than TR80 (27.
49-5 1). Therefore. il was of great inleresl 10 determine
whether the above inhibitors could be used \0 dist ingu ish
between the signals used by the two dist inct TNFRs in
KYM-I cells. Receptor-specific agonistic Abs were used
to selective ly induce cytolysis in KYM- l ce ll s. In a first
set of experiments, inhibitors were used at a cons tant subtoxic concentralion and the respective TNFR sti muli were
titrated. The data oblll ined (Fig. 4) indeed revea l a different ial sensitivity to the inhibitors and suggest a segrega-
tion in the signal pathways used by TR60 and T R80 for
induct ion of apoptosis in KYM - l cells. Thus, TR60-triggered cytotox icity could be antagonized by pargyline and
H8 (Fig. 4, A and B). but nOt by NDGA ( Fig. 4C). In
contrast, the inhibitor NDGA specificall y interfered with
TR80-l1lediated cyto lysis induced by a TR80-speciHc scrum (Fig. 4G), but no inhibition cou ld be obwi ned with
either H8 or pargyline (Fig. 4 , E and F). To ensure that the
observed se lectiv ity of H8, pargyline , and NDGA WllS not
concentrat ion dependent, we titrated the inhib itors in a
second set of experiments. Representative results are
The Journal of Immunology
.
:
:
'\oO~ • • •
FIGURE 2. Kinetics of DNA fragmentation on selective
TNFR activation. KYM-l cells were treated as described in
Figure I with TNF (A ), M80 serum (8 ), or H398 plus goal
anti-mouse ISG (e) for the indicated times. Control lanes (cl
represent 18-h stimulation with medium (/\ ), control rabbit
seru m (8), or goa t anti-mouse ISG .110ne (C). DNA was extracted and subjected to electrophoresis on a 1.1% agarose
gel in the presence of elhidium bromide.
shown in Figure 5. which clearly demonstrates that
the selectivity of the inhibilOrs was independent of
the concentration used. Again, only p3rtia l protection
could be obtained because of the intrinsic loxicity of these
compou nds.
Because ill KYM-l ce lls the selective activ3tion of each
of the TNFRs fin all y leads to the development of apoptosis. il seems likely thtH bolh signal pathways me rge at a
point before the relevant intrilccllular effector molecules of
apoptosis arc act ivated. Indeed. wc found that down-regulation of ROI formation by amy tal sign ificant ly amel iorated both TR60- and TRSO- induccd cylOlysis (Figs. 4. f)
and H and 5D}. Taking into consideration that amy tal possesses a consi derable intrinsic cytotoxici ty at the concentralions used. TR60-triggered effects could be blocked approximately 30% (Fig. 4D: O.S J-Lg/ml H39S mAb) and
T RSO-med iated effects were inhibited to a lesser. bUI significant. extenl (20%: Fig. 4H: seru m dilution I :SOO). On
Ihe basis of these resu lls wc conclude Ihat. in KYM -I
ce lls_ the signal s induced by both receptors mighl be connected to Ihe mi tochondrial ROI system.
Wc howe recently shown that induct ion of TNF resistance in KYM - I ce lls affects both TNFRs (52). Provided
the sigll<lling cascades triggered by the twO TNFRs segregate. it wou ld be possible to establish selectively resislant
ce ll lines derived from KYM · 1 by prolonged specific st imulation of on ly one of the two TNFRs. Indeed. after 2 to 4
wk of cult ure in the presence of Ihe TR60-specific agonistic mAb, h1 r- l. the surviving ce lls had become highly
1967
resistant toward a subsequent cha llenge with the same
stimulus ( Fig. 6A). Remarkably. Ihe sensitivity of these
cell lines, designated K YM-60res, toward MSO-induced
apoptosis was only slight ly reduced (Fig. 68). The reciproca l resu lts were obtained uCier a se lectio n with MSO
serum. In this case, a strong desensit izat ion o f TRSO signaling (Fig. 68) hardly affected se nsitivity toward TR60induced cytotoxici ty (Fig. 6A).
Next, wc wamed 10 know the effects of T R60 or TR SD
desensitizat ion on the expression of TNFRs by ligand
compclition studies with iodinated TNF and receptor specific Abs. As published previously (52), neither stimulus
had any effects on TR60 cel l surfac e ex pression, whereas
TNF or MSO treatment led 10 a reduction in TRSO receplor
expression. When TRSD-resistam ce ll lines had been established after approximate ly 14 days of continuous sti mulation, these cells expressed 3000 to 6000 TRSO mo lecules/cell. On additional culture in the presence of TRSO
agonists, this number furt her declined to levels below detection « I00/cc lJ ) within 2 to 3 mo (52: and data not
shown). However, induction of resislance in the KYMSOres cells s hown in Figure 6 c:tOnot be attributed to homologous reeeptor down-regulation. because these cells still
expressed 4000 TRSO molecules/cell (data not shown). l1lis
suggCStS a postreccptor level of resislance induct ion prccceding complete loss of membrane receptors. in contrast. KYM60rcs cells, having gained strong resistance toward TR60specific agonists (Fig. M), showed no change in either TR60
or TRSO membrane expression. l1lese data indicate that the
sensitivity changes in this subline arc strictly confined to a
postreceptor level , i.e .. afrect signal transduction.
Although the deve lopment o f selective receplor resistance was reproducible in a number of independent experiments. it was not possible to gener:lle ce ll lines with stable
phenotypes. First. the KYM-60res line cou ld be revened
readily to higher sensitiv ity within I wk after ,Ibrogat ion
of the receptor-specific stimulus. Second. with a prolonged
period (> 2 mol of constant selecti ve pressure, both cell
lines. KYM-60res and KYM -SOres. converted 10 a crossresistant phe notype in which both T NFR signal pathways
were lIffected (d'lliI not shown).
Discussion
TIle role of the two defined TNF membrane receplors is
currelllly discussed in a controversial manner. In the majority of experimental systems. the abilit), of TNF to exe rt
cytotox ic effe cts has been clearly attributed 10 TR60 (12.
34, 36, 37). This investigat ion , however, unilmbiguously
demonstrates that TR SD is per se able to trigger apoptosis,
i.e., programm ed cell death. TypiCll1 symptoms of apop+
tosis. like DNA fragmentation and chromatin condensation at the nuclear membrane. were noted on independent
triggering of each of the two TNFRs (Figs. I and 2). A
INDUCTION OF APOPTOSIS VIA 80TH TNF RECEPTORS
1968
iI
0
....
.'"
~
eo
~
~
.~
~
8
z
................................................. -
................................. ... .......................... ...
-
"
,.,
.
'G- ,'"
,
,.,
•
•••••••••••••
• _ _"
••••••••••• •
_
.................. ... ...... ... . . . . . . . . . h
'_'
........................
-
FIGURE 3. Influence of various inhibitors of intracellular signal pathways on TNF-mediated cytotoxicity. KYM-l cells were
cultured for 18 h in microtiter plates with or without 250 p&,ml TNF (approximately EDso) in the absence or presence of
genistein (3.2 ~) , Ha (15 pM), H89 (0.5 IJM), H7 (6 .3 pM), K-252a (250 nM), K-252b (2 .5 ~), staurosporine (25 nM),
calphostin C (50 nM), neomycin (1 mM), NOGA (1 0 IJM), quinacrine (2.5 pM), NCOC (50 JLM), BPB (2.5 I'M), 0609 (120 pM),
liCl (2 .5 mM), pargyline (1 .25 mM), amytal (400 IJM), TIfA (45 pM), or antimycin (5 pM). The maximum concentration for
each drug that did not, on its own, exhibit more than 10% cytotoxic effect over the assay period was determined and chosen
for the experiments. Cell viability was determined by formaz an production and evaluated in an EUSA reader at S60 nm . Shown
is the influence of the drugs on TNF-induced cytotoxicity setting the TNF effect without the drug as 100% and the OD in the
presence of the inhibitor alone as 0%. Experiments were performed in triplicate and the mean SO of at least three independent
experiments is presented (signifi cance was evaluated by using Student's Hest:·p < 0.05; up < 0.01).
minor difference observed was the slower kinetics of ap·
oplotic development after stimulation ofTRSO when corn·
pared with those of selective TR60 stimulation or of combined receplor triggering. Significant DNA fragmentation
could be detected after 4 h of stimulation with TNF.
whereas 8 hand 12 to 18 h of TR60 and TR8D stimulation,
respectively. were necessary 10 obtain comparable effects
(Fig. 2). These differential kinetics in apoplotic development, confirmed by visual microscopic controls. could ei·
ther reflect differences in signal transduction pathways or
be a rcsull of the different reccplor cross·linking efficiencies by the Abs vs TNF. In addition, because antagonistic
Abs specific for each of the two receptors do inhibit TNF
cytotoxicity in KYM· I cells (40). cooperative effects of
TR60 and TR80 on TNF triggering are conceivable.
Because KYM·l cells express extraordinarily high numbers ofTRSO compared with norma1tissucs the ability of this
receplor to induce programmed cell death could be attributed
to this fact and, consequently, could represent a specific property of oruy this particular cell line. However, investigations
performed with the myeloid granulocyte-macrophage CSFdependent cell line, GMSO (53), revealed that overexpres·
sion of TRSO is not a prerequisite to function as an inducer of
apoptosis. GMSO cells express TR80 at levels comparable 10
primary cells (- 2000/cell) and also develop apoptosis on se-
iective stimulation of TRSO (S. 6z, manuscript in prepara·
tion). Therefore, it seems conceivable that TRBO receptors
could be involved in TNF-mediated apoptosis in biologically
relevant situations. e.g., in negative selection of thymocytes.
The recent work of Hemandez-Caselles and Stutman (6) is in
accordance with this view. These authors suggested that
TROO is involved in both growth stimulation and induClion of
apoptosis by TNF. dependent on the culture conditions used.
The availability of a cell line that shows an identical
cellular response on selective stimulation of each of the
two TNFRs prompted us to study identity or differences in
the intracellular signals involved. In a first step, we tested
a broad range of inhibitors of potential intracellular sig·
naling molecules for their effects on TNF-induced apoptosis in KYM-l cells. From the emerging pattern of mod·
ulalion of TNF responsiveness, the following conclusions
can be drawn. A number of PKC inhibitors were found to
act in an additive/synergistic way with TNF, suggesting
that they either enhance TNF-mediated cytotoxicity directly or block concomitantly induced protection mechanisms. These enhancing compounds include H7, slaurosporine, and K252A and B, substances that have been often
used as inhibitors of PKC. but certainly have a broader
target spectrum (32, 54). On the olher hand, calphoslin C,
currently regarded as a highly potent and specific inhibitor
The Journal of Immunology
1969
.,
tie (I'M)
PqyIine (mM)
D
D
50
D
"
~~A
.~
~
50
".
50
G
50
H
o
10'"
1~
1~
M80 aon.m (dihAion)
FIGURE 4. Inhibition of receplor-specific induction of cytotoxicity. KYM-l cells were incubated with increasing concentrations of either TR60-specific mAb H398 plus 3 J..lg/ml
goat anti-mouse rgG (A through U) or TRaO-specific rabbit
serum (E Ihrough H) in the absence (O) or presence (.) of
1.25 mM pargyl ine (A and E), 15 pM Ha (B and F), 10 J.!M
NDCA (Cand G), orO.5 mM amylal (Dand H). Viability was
determined as described in Figure 3, The maximum cytotoxic
effect of the respective agonist under the given experimental
conditions was 100% and the medium control, 00/0, i.e., the
intrinsic cytotoxic effect of the respective inhibilor was not
subtracted. A representative experiment is shown .
of several PKC isoenzymes. did not affect TNF-triggered
cytolysis (Fig. 3). These results question the role of calphostin C-sensitive PKC isoenzymes in signaling of. and
in protection from, TNF cytotoxicity in KYM-l cells (26,
32), altbougb the role of serine/threonine-specific PKs in
TNF signal transduction seems to be evident (2),
Quite remarkable protective effects were found with an·
other kinase inhibitor, the isoquinoline sulfonamide derivative H8. Although often used as a PKA-specific reagent,
this inhibitor is now also known 10 affect various other
kinases, depending on the concentration used (54). Morc·
over, specific involvement of PKA in TNF signa ling
seems unlikely because the related inhibitor H89 was totally inefficient. This derivative has a very similar structure
to H8 but has a higher affinity for cAMP-dependent kinases (IC~ of 0.048 vs 0.48 p.M) and is, thus, expected to
block PKA more effectively (55; and information of the
supplier). Taken together, the data obtained with the PK
inhibitors indicate a role of PKs in TNF signaling and
./
•
~
o
~.
~
50
"
•
~~
•
"
D
".
"
D
~
., '-'.D
D
D
50
'00
Am)'l.alCmMJ
FIGURE 5. Dose-response c:urves of inhibitors, KYM-1
cells were treated with serial dilution of inhibitors alone (0 )
or with 0.75 J..ls'ml H398 plus 3 ",glml goat anti·mouse IgG
(0) or MBO serum (1 :1500; . ) in the presence of serial diluted inhibitors. Viability was determined as described in Figure 3. The maximum cytotoxic effect of the respective agonist
under the given experimental conditions was 100% and the
medium control. 0%. Representative experiments are shown.
l
.
D
~ "
1
OD
OD
'00
•
••
A
.......
~,
..
• """",,. ,... , ,.'
D
mAb htrl
(d~ution)
•
•
."
D
eo
eo
D
,...
,.., ,,.,
'00
MaO lerum (dijution)
FIGURE 6. Development of receptor-specific resistance in
KYM-l-related cell lines. KYM-l cells had been cultured in
the presence of increasing concentrations of TR60· (KYM60res cells) or TRao- (KYM-aOres cells) specific agonists.
Typical sensitivity patterns after 3 wk of culture are displayed
in comparison with untreated KYM· 1. Cells were washed extensively then stimulated with the TR60·specific mAb htr-1
(A) or MaO serum (8) for la h, and viability was determined
as described in Figure 3. The maximum cytotoxic effect of the
respective agonist under the given experimental conditions
was 100%. A representative experiment is shown.
TNF-induced protective mechanisms but suggest that nei·
ther conventional PKC sUbtypes nor PKA are critically
involved in either process in KYM· t cells.
Furthermore, the use of various inhibitors or phospholipid metabolism suggested involvement of the PLA2"Ii·
poxygenase pathway in TNF-mediated cytotoxicity, which
is in full agreement with data obtained by other investi·
gators (22-24). The experimental results obtained in this
study, however, did not reveal exact components involved
1970
in this process because the PLAz inhibitors SPB and quinacrine, in contrast with neomycin and NDGA, were inef-
fective. Regardless of the precise location of inhibitory
action, the identification of inhibitors that interfere with
TNF-induced cytotoxicity in KYM-l enabled us to investigate whether TNFR signaling pathways segregate. Indeed. some of the inhibitory compounds revealed a clear
segregation in this regard. The HS-sensitive PKs that are
involved in TNF-mediated cytotoxicity can be clearly
linked 10 TR60-induced pathways (Fig. 4). On the other
hand, NDGA, often used as an inhibitor of Iipoxygcnase
pathway, but also known as an inhibitor of mono-oxygenase and as a radical scavenger (56), selectively inhibited
TRBO-mediated apoplosis (Fig. 4). Accordingly, lNF-mediated apoptosis in KYM-l cells can be initiated by both
TNFRs using at least initially distinct signal pathways.
These two pathways apparently function in an additive
manner, as demonstrated previously (40), and most likely
merge at some point upstream of mitochondrial ROI production before the activation of the apoptotic effector mol.
eculcs. This is suggested from data obtained with the ROI
production inhibitor amy tal, which reduced cytotoxicity
induced by both TROO and TR80 (Fig. 4),
KYM-I cells have been shown to readily develop lNF
resistance on prolonged treatment with TNF (52), The differential sensitivity of TROO- and TR80-mediated apopto-sis 10 various inhibitors of the signal transduction mechanisms raised the possibilily 10 selectively induce resistance
at a given receplor sUbtype. Cell lines developing such a
receptor-specific resistance were, in fact, reproducibly obtained and, thus, further support the view of a distinct signal transduction pathway usage of the two TNFRs. Together, these data suggest that for both TNFRs response
limiting steps in the respective signaling pathways must
exist which 1) are distinct from each other and 2} can be
regulated independently. On the other hand, a limited
cross-desensitization was reproducibly found in all 60res
?nd SOres cell lines (Fig. 6), indicaling that some overlap
10 TROO and TR80 action must exist. This could be effec.
live al the slage of generation of apoplotic effector molecules and/or directly upstream at a late step in the intracellular signal pathway thal is shared by both receplors.
In the case of KYM-80res cells, an additional mechanism mighl be involved in the devclopment of resistance
because these cells reduced TR80 membrane receptor expression (52; data not shown). The importance of this
gradual membrane receptor loss for generation of resistance is unclear at present. For example, we noted that
TNF resistance preeeeded complete loss of receptors, and
~lIy resistanl lines were Oblained that still expressed sigmficant numbers of TR80 membrane receplors. Although
the mechanism of receptor down-regulation has not been
addr~sscd in this study, Ab-induced receptor shedding is
poSSible, because TR80-spccific mRNA levels remained
unchanged (data nOI shown), In any case, TR80 down-
INDUCTION Of APOPTOSIS VIA BOTH TNf RECEPTORS
regulation is induced by TR80 activation, itself, and does
1'101 represent cross-modulation triggered by TR60.
Currently. the majority of TNF responses has been attributed to TR60, whereas the overall contribution and in
particular the separate functional role of TRSO in TNF signaling was unknown. Nevertheless. in a number of di1ferent
cellular systems, cooperative effects of TR60 and TRSO have
been described, These include cytotoxic effects of lNF on
U937 cens (38), stimulatory activity for T cells and thymocytes (57), and enhancement of surface Ag expression in co-.
Ion carcinoma cells (14). We could now show that TRSO is
involved in lNF-medialed signaling and present evidence for
the existence of receplor selective signal pathways. These
pathways may lead to the same cellular response yet arc subjected to largely independent cellular control mechanisms,
This finding suggests that it might be possible to develop
strategies that circumvent the usually rapid process of lNF
desensitization, Accordingly, an alternative shorHerm Ireatment with receptor.specific agonists could prevent a general
TNF unresponsiveness and, therefore, could be superior to a
treatment with the natural ligand that affects both receptors
simultaneously.
Acknowledgments
We thank Dr. M. Brockhaus, Hofl"mano·u Roche, Basel, Switzerland,
for providing hlr· 1 Ab and Dr. I.-M. von Brocn, Knoll AG, Ludwigshafen, Germany, (or r'1"NF. The e:c.cellent technicalassislance of Beale
Maxeiner and Beate Rehkopf is highly appreciated.
References
I. Jlluela, M. 1991. Biologic activities and mechanisms of action of
tumor necrosis factor-aJcachectin. Lab. Im·est. 64:724.
2. Adolf, R. A, M. Grell, and P. Scheurich. 1994. TUrnor necrosis
factor. In Epidermnl Growth FOCrQTS and Cytakines. T. A Luger and
T. Schwan, eds. Matcd Dckker, Inc., New York, p. 63.
3. Kronke, M., S. Schiltze, P. Scheurich, and K. Pfizenmaicr. 1992.
TNF signal tronsduaion and TNF· responsive genes. Immunol. SeT.
56: /89.
4. Ru~ero, ~.: K. Latham, and C. Bagtioni. 1987. Cytostatic and cy.
totOXIC actIvity of lumor necrosis factor on human cancer cells.
J . ImmunaL JJ8:271/.
S. Meager, A. 1991. A cytotoxicity assay for tumour necrosis factor
using a human rhabdomyosarcoma cell line. J. Immunol. Methods
144:14/.
6. Hernandcz-Caselles. T., and O. Stulman. 1993. Immune runctions of
~F. I. TN~ i~~cc.s I~osis of mouse thymocytes and can also
stimulate or mhlblttlAt-lnduccd prolifenllion depending OD the m n.
cent rdtion of mitogenic costimulation. J. ImnultIol. 15/:3999.
7. Loetscber,. H., Y. C. Pan, H. W. Lahm, R. Gentz. M. Brockhaus,
~ . Tabuchl, and W. Lcsslauer. 1990. Molecular cloning and eltpres.
Slon of tbe human SS· kd turnor necrosis factor rcceplor. Cell 6/:35 / .
S. Schalt, T. J., M. Lewis, K. J. Koller, A. Lee, G. C. Rice, G. H. Wong,
T. Gatanaga, G. A. Granger, R. LenlZ, H. Raab, W. J. Kohr, and
O. V. Gocddel. 1990. Molecular cloning and expression of a receptor
for hum an tumor necrosis factor, CeIl6J:J61.
9. Smith, C. A., T. Davis, O. Anderson, L Solam, M. P. Beckrnann,
R. l eny, S. K. Dower, D. Cosman, and R. G. Goodwin. 1990. A
receplor (o~ turnor n.ecrosis factor defines an unusual family of cellular and Viral protems. Science 148: 1019.
10. Ilimmlcr, A , I. Maurer-Fogy, M. Kronke, P. Scheurich, K. PflZenmaier,
M. LanIZ, I. Olsson, R. Hauptmann, C. Stratowa, and G. R. Adolf. 1990.
The Journal of Immunology
Molecular cloning and expression of human aDd rat tumor necrosis be.
tor rcccptor chain (p6O) and its soluble deriva1ive tumor ncaosis (lIdor.
binding protein. DNA. Ct/1. Biol. 9:705.
11 . Knmm~r, ~. H., and f.<..m. Dcbatin. 1992. Autoimmunity: when
apoptoslS falls. CUFf. BIOI. 20'18J.
12. Thoma, B., M. G rell. K.. pfizcnmai<:r. and P. Scheuricb. 1990. Identification of a 6O-kD tumor neaosis factor (TNF) rcccptor as the
major sign al transducing component in TNF responses. J. Exp. Med.
I n:1019.
13. Ware, C. F., P. D. Crowe, T. L Van Arsdale, J. L Andrews, M. H.
Grayson, R. Jen:y. C. A. Smith, and R. G. Goodwin. 1991. Tumor
necrosis factor (lNF) rec:eptor expression in T Iympbocytes: differ.
ential regu latio n of the type I TNF receptor during activation of
resting and elTector T cells. J. Immunol. 147:4229.
14. Scbcurich, P., M . Grell, A. Meage r, and K.. Pfizcomaier. 1993. Agonistic and antagonistic an tibodies as a 1001 In study the functional
rol e of human tumor necrosis factor rcccptors. In Tumor Necrosis
Focror, Vol. 4. W . Fiers and W . A. Buurman, «Is. Karger, Basel,
Switzerland, p. 52.
15. Brockhaus, M., 11 . J. Schoenfeld, E. J. Schlaege r. W. Hunriker,
W. Lcsslauer, and H. Loetsche r. 1990. Identification of two types of
turnor necrosis factor receptors on human o::U lines by monoclonal
antibodies. Prrx:. NOli. Mad. Sd. USA. 87:1127.
16. Wright, S. C., and 8. Bonav ida. 1987. Studies on tbe mechanism of
11.
18.
19.
20.
21.
natural killer cell·mediated cytotoxicity. VII. Functional comparison
of hu man naturalltiller cytotoxic factors with rccombinant Iympbotoxin and tumor necrosis factor. J. ImnIl4nof. /38:1791.
Dell, C. A., M . Gatanaga, E. K. Ininns, F. Cappuccini, R. S.
Yamamolo, G. A. Grange r, and T. Gatanaga.. 1991. Enhancement of
Iympbokinc·activated T killer cell tUIDOr necrosis factor reccptor
mRNA transcription, tumor necrosis factor receptor membrane ex·
pression. and tumor necrosis factor/lympbotoxin release by 11... 1/3
llA, and 1L-6 in vitro. J. Immunol. 146: 1522.
Perez, C., I. A1bcr1. K. DeFay, N. Zachariades, L.. Gooding, and
M. Kriegler. 1990. A nonsccretable cell surface mutant of tu mor
neClOSis factor (TNf) kills hy cell-to-ccll co ntact. Cell 63:251.
Giroir, B. P., T. Brown, and B. Bcutler. 1992. Constitutive synthesis
of tumor necrosis factor in th e thymus. Proc. Narl. Acad. Sd. USA
89:4864.
lalTick, J. W., and S. C. Wright. 1990. Cytntoxic mechanism of
tum or necrosis factor-a. FASEB J. 4:1215.
Laster. S. M., J. G. Wood, and L. R. Goading. 1988. Turnor necrosis
factor can induce both apoptotic and necrotic forms of cell lysis.
J. lmmunol. 141:2629.
22. Sulfys, P.• R. Beyaen, F. Van Roy, and W. Fiers. 1987. Reduced
lumour necrosis faelor· induced cytotoxicity by inhibitors of the
arachidonic acid metabolism. Bioch~m. Biophys. Res. Commun. 149:
715.
23. Chang. D. J., G. M. Ringold. and R. A Helier. 1992. Cell kil ling and
induction of manganous superoxide dismutasc by turnor necrosis
factor· o is mediated by lipoxygena.sc mctabolites of arachidonic
acid. Biocltem. Biophys. Rn. Commun. 188:518.
24. De Valck, D., R. Beyaen, F. Van Roy, and W. Fiers. 1993. Tumor
necrosis factor cytotoxici ty is associated with phospholipasc- D activalion. Eur. J. Biochem. 212:491.
25. Schiirze, S., O. Bcrkovic, O. Tomsing, C. Unge r, and M. Krooke.
1991 . Tumor necTosis factor induces rapid production of I ' 2'diacyl.
glycero l by a phosphatidy1choline·specific phosptlolipasc C. J. up.
Med. 174:971
26. Deyaen. R. , B. Vanhaesebrocck, K. Heyninck. E. Boone, D. Dc
Vatck. K. Schulu·Qsthofl'. G. Hacgeman, F. Van Ray, and W. Fiers.
1993. Sensitization of tumor cells 10 tumor necrosis faClor acti on by
the protein kinase inhibitor staurosporine. Cancer Res. 51:2623.
27. Schu tzc, S., K. Pollhofl', T. Machleidt, D. Bcrkovic, K. Wi egmllD n,
and M. Kt6oke. 1992. TNF activates NF-I<B by phosphatidylchoJinespecific phospholipase.C.induced acidic sphingomye lin breakdown.
Cell 71:765.
28. Schulze.Osthoff, K., A. C. Biller, B. Vanhacsc:brocck, R. Deyaclt,
W. A.. Jacob. and W. Fiers. 1992. Cytotoxic activity of tumor ne·
crosis factor is mediated by early damage of mitochondrial func-
1971
tions: evidence for the involvement of milocbondrial radical generation_ J. Bioi. Chem. 267:5]17.
29. KOnig, M., O. Wal1 ach, K. Resch, and H. Holtmann. 1991.induelion
of hyporcsponsiveness to an early post-binding elTect of tumor necrosis fador by tumor necrosis factor itself and interleukin I . Eur.
J. Immunol. 21:1741.
30. Wong, G. H., J. H. Elwell, L W . Obcrley. and D. V. Gocddcl. 1989.
Manganous suptroxidc dismutase is essential for cellular resisance
10 cytotoxicity of turnor necrosis factor. Cell 58:923.
31. IUttel1, M., lDd D. Wissing. 1993. Heat·shock proteins proted cells
from monocyte cylolox.icily: possible mechanism of self-protection.
J. up. Med. 177:211.
32. Hamamoto, Y .• T. Matsuyama, N. Yamamoto, and N. Kobayashi.
1990. Augmentation of cytotoxic efl'ect of tumor necrosis facto r on
human immunodeficiency virus-infected cells by staurosporinc, a
potent protein kinase C inhibitor. Cancer Res. 50:5287.
33. Sampson, L E., A Mire-Sluis, and A. Meager. 1993. Protein kinase
C-depcndenl phosphorylation is involved in resislance to tumour nectosis factor-Q-induced cytotoxicity in a human mo nocytoid cell
line. Biochem. J. 292:289.
34. Engc1mann, H., H. Holtmann. C. Brakebusch. Y. S. Avni, I. Sarov,
Y. Nopbar, E. Hadas, O. Leitner, and D. Wallach. 1990. Antibodies
to a sol uble form of a tumor necrosis factor (TNf) reccptor have
JNF·like activity. J. BioI. Chem. 265:14497.
35. Lcwis. M., L A.. Tanaglia, A.. Lee, O. L Bcnnett, G. C. Rice, G. H.
Wons. E. Y. Chen, and D. V. Gocddcl. 1991. Ooning and expres·
sion of cONAs for two distinct murine tumor necrosis factor rec:eptors deulOlUtrate ooe re«plOt' is SpcciC3 specific. Proc. NOlL A.:<Id.
Sei. USA 88:2830.
36. Tanaglia, 1.- A., R. F. Weber, I. S. Figari, C. Reynolds, M. A Pal·
ladine, Jr., and D. V. GoeddeJ. 1991. The two dif('erent rec:eptors for
[\lmOr necrosis factor mediate distinct cellular responses. Proc. NaIl.
Acad. Sci. USA 88:9292.
37. Tanaglia, L A.. T . M. Ayrcs, G. H. W. Woo&. and O. V. Goeddel.
1993. A novel domain w ithin the 55·kd TNF receptor signals cel!
death. Cdl 74:845.
38. Shalaby, M. R., A Sundan, H. LoclSChcr, M. Brockhaus, W. L.c:sslauer,
and T. Espc:vik. 1990. Binding and reguJatioo of cellular functions by
monoclonal antibodies against human turnor necrosis factor receptors.
J. £Jp. Med 172:1517.
39. Helier, R. A., K. Song, N. Fan, and O. J. Chaog. 1992. The p70
wmor necrosis factor receptor mediates cytotoxicity. Cell 70:47.
40. Gre ll, M., P. Scheurich, A. Meager, and K.. pfizenmaier. 1993. TR60
and TRSO turnor necrosis factor (fNF}rcceptors can independen tl y
mediate cytolysis. Lymphokin~ Cyrokine Res. 12:141.
41. Sekiguchi, M., Y. Shiroko, T. Suzulti, M. (mada, M. Miyahara, and
G. Fujii. 1985. Characterization of a human rhabdomyosarcoma cell
strain in tissue culture. Biomtd. Pharrrt.aCOlher. 39:372.
42. Bruns, R. F.• F. D. Miller, R. L Merriman, J. J. Howbc rt, W. F.
Heath, E. Kobayashi. I. Takahashi, T. Taroaoki, and H. Nakano.
1991. Inhibition of protein kinase C by caJpbostin C is light depen.
de nt. Biochtm. Biophys. R~$. Commun. 176:288.
43. Wyllie, A H., J. F. R. Kerr. and A. R. Cunie. 1980. Cell death: the
significance of apoptosis. Int. Rev. Cyrol.68:251.
44. Wyllic, A. H. 1980. Glucocorticoid·induced thymocyte apoptosis is
associated with endogenous endonuclease activation. Nature 284:
551
45. Mallhew5, N., M. L Nea le, S. K.. JacOOn, and J. M. Stark. 1987.
Tumour cell killing by tumour necrosis factor: inhibition by anaer·
obic condi tions, free·radical scavengers and inhibilOrs of arachidonate metabo lism. Immun ology 62: 151.
46. Schulze-Osthofl', K., R. Bcyaert, V. Vandcvoorde, G. Ha cgeman,
and W. Fiers. 1993. Depletion of the mitocbondrial electron transpon
abrogatcsthe cy totoxic and gene-inductive effects ofTNF. EMBO J.
12:3095.
47. Unglauh, R., B. Maxeiner, B. '11KIrna, K. Pfizenm aier, and
P. Scheurich. 1987. Down-regulation of tumor necrosis faclor (TNF)
sc nsitivity via modu lation ofTNF binding capaci ty by protein kinase
C activators. J. up. Med. /66:/788.
48. Scheurich, P., G. KObrich, and K.. PfizenmaieT. 1989. Antagonistic
conlInl of turnor necrosis factor rec:eptors by protein kinascs A and
1972
c: enh ance men! of TNF recep!or synthesis b)' protein kinase A and
ITansmodulation of receplor~ by prOlein kinase C. 1. £rp. /lft'd J70:
9-/7.
49. Wiegmann. K.. S. Schiltle. E. Kampcn, A Himm1cr. T. MscIlleidl. and
M. KrUnke. 1991. Hum an 55-kDa rcttpIor for Moor necrosis factor
ooupled 10 signal transduction casc-Jdcs. 1. Bio/. CIU'IIL 167: 17997.
50. Yanaga. F.. and S. P. Wa tson. 1991. Tumor necrosis factor-a s!imu13!es sphingom)elinase through the 55-I; Oa receptor in HL-60 cells.
PEBS UTI. 3N:197.
5 1. Wong. G. H.. LA Tartaglib. M. S. lee. and O. V. Gocddcl. 1992.
Anth iflll aClhil~ of tumor necrosi~ factor is signaled through the
55-k03 t)PC 1 TNF roceplOr. J. hlllIllIllQI. N9;J350.
52. Meagt'r. A.. L E. Sam pMln. M. Grell. and P. Scheurich. 1994. IX\ elopment of resistant't to tumour necrosis factor (fNF) in KYM -1
cells 3ffects both TNF receptors. C.I/okim· 5:564.
53. OZ. S.. H. Tinelbach. R. Fahsold. R. Schaetzl. C. l3uhrer. 1. I\ v.podicn.
and J. R. !\.alden. 1990. E.sIablishmcnt and characteriLation of a gran-
INO UCTION OF APOPTOSIS VIA BOTH TNF RECE PTORS
uJocyte-macrophage coioo)"-SlimuJating factor-depc:ndcm human myeloid ce ll line. 81()()(176:578.
5-1. Casnellic. 1. E. 1991. !'rotein kinase in hi bitors: probes for the fune_
tions of prolein phosphorylation. '\ilL Phimll(i("O/. 22:167.
55. IHdakD. 11.. M. Jnagald. S. Kl1\\ amOlO. and Y. Sasaki. t9S-!. hQquinolinC!.ulfonamidcs. no\ el and potenl inhibitor, of c)dic nud eotiUe-dependl·nt protein ki nase and protein kinase C. lJIocllt'llU.\/n
!J:5036.
56. Agarv.·al. R.. Z. Y. Wang. I). P. Btk. and H. " Iukhtar. 19'J1. Nordih)droguaiaretic acid. an inhibitor or lipox}"genasc. atw inhibits c)lochrome P-450-medialcd monoo.xyge na...c acth ity in ral epidermal
and he~tie microsomes. Drus: Mt'/llb. Dlspos. 19:620.
57. Gehr. G .. R. Genll_ 1\1. Brocthau~. H. Loctscher. and W. Les~lauer.
1991. Dolh lumor necrosis foctor receplor type:. mediate proli(era_
ti\c ~isna" in human mononuclear cell aelivation. J Immllllfll. f.J9.
911