Tumor necrosis factor-a converting enzyme is processed
by proprotein-convertases to its mature form which is degraded
upon phorbol ester stimulation
Kristina Endres, Andreas Anders, Elzbieta Kojro, Sandra Gilbert, Falk Fahrenholz and Rolf Postina
Institute of Biochemistry, Johannes Gutenberg-University, Mainz, Germany
Tumor necrosis factor-a converting enzyme (TACE or
ADAM17) is a member of the ADAM (a disintegrin and
metalloproteinase) family of type I membrane proteins and
mediates the ectodomain shedding of various membrane-
anchored signaling and adhesion proteins. TACE is syn-
thesized as an inactive zymogen, which is subsequently
proteolytically processed to the catalytically active form. We
have identified the proprotein-convertases PC7 and furin
to be involved in maturation of TACE. This maturation
is negatively influenced by the phorbol ester phorbol-12-
myristate-13-acetate (PMA), which decreases the cellular
amount of the mature form of TACE in PMA-treated
HEK293 and SH-SY5Y cells. Furthermore, we found that
stimulation of protein kinase C or protein kinase A signaling
pathways did not influence long-term degradation of mature
TACE. Interestingly, PMA treatment of furin-deficient
LoVo cells did not affect the degradation of mature TACE.
By examination of furin reconstituted LoVo cells we were
able to exclude the possibility that PMA modulates furin
activity. Moreover, the PMA dependent decrease of the
mature enzyme form is specific for TACE, as the amount of
mature ADAM10 was unaffected in PMA-treated HEK293
and SH-SY5Y cells. Our results indicate that the activation
of TACE by the proprotein-convertases PC7 and furin is
very similar to the maturation of ADAM10 although there is
a significant difference in the cellular stability of the mature

at a characteristic motif, which is located between the
prodomain and the metalloproteinase domain. Proprotein-
convertases form a family of calcium-dependent endopro-
teinases, which presently comprises seven distinct members,
including furin, PC2, PC1/PC3, PACE4, PC4, PC5/PC6
and PC7/PC8/LPC [16,17]. A large number of proproteins
with various specificities are processed by these subtilisin-
like convertases. Typically, cleavage occurs C-terminal to
the common consensus sequence RX(K/R)R. Proteolytic
activation of substrates, mediated by PC7 or furin, takes
place in the trans-Golgi network, in endosomes and at the
cell surface [18,19]. Both convertases share an overlapping
substrate specificity and therefore the selectivity of substrate
proteolysis depends on each ones exact cellular localization.
As intracellular trafficking is regulated by their cytosolic
domains, which contain different sorting motifs, it is
possible that the localization of PC7 is distinct from that
of furin [19]. Recently, we have demonstrated that over-
expression of the proprotein-convertases PC7 and furin in
Correspondence to R. Postina or F. Fahrenholz, Institute of
Biochemistry, Johannes Gutenberg-University, Becherweg 30,
D-55099 Mainz, Germany.
Fax: + 49 6131 3925348, Tel.: + 49 6131 3925833,
E-mail:
Abbreviations:Ab, amyloid b peptide; ADAM, a disintegrin and
metalloproteinase; APP, amyloid precursor protein; APPsa, a-secre-
tase cleaved soluble APP; APPsb, b-secretase cleaved soluble APP;
DMEM, Dulbecco’s modified Eagle’s medium; HEK293, human
embryonic kidney cells; PC, proprotein-convertase; PMA, phorbol-
12-myristate-13-acetate; TACE, tumor necrosis factor-a converting

has been reported to increase the basal and protein kinase C-
dependent APPsa release [28], but the purified enzyme failed
to cleave a synthetic peptide at the major a-secretase
cleavage-site [29]. In contrast, ADAM10 has been found to
have constitutive and regulated a-secretase activity as well as
many other properties expected for an a-secretase [27].
Additionally, in situ hybridization analysis in human
cortical neurons provided evidence for the coexpression of
APP with ADAM10 and b-site APP-cleaving enzyme
(BACE)suggestingthatADAM10ismostlikelythe
physiologically relevant a-secretase [30]. Finally, experi-
ments performed with TACE-deficient cells pointed to a
participation of TACE in only the regulated, protein
kinase C-stimulated a-secretase pathway [26,31]. In another
cellular context a constitutive a-secretase activity of TACE
was demonstrated [32].
As PC7 and furin act as pro-a-secretase converting
enzymes [20], we investigated the proteolytic processing of
TACE by overexpression of these convertases in HEK293
cells. We were able to show, that both proprotein-conver-
tases contribute to TACE maturation. Moreover, we
examined the effect of PMA on the processing of endo-
genous TACE and ADAM10 in various mammalian cells
and discovered a reduction in the amount of mature TACE
compared to ADAM10 after PMA treatment.
Experimental procedures
Primary antibodies
The following antibodies were used: anti-ADAM10, a
polyclonal rabbit antibody against endogenous ADAM10
and anti-TACE, and a polyclonal rabbit antibody against

HEK293 cells were cultured in the presence of 30 l
M
decanoyl-RVKR-chloromethylketone (Bachem AG, Swit-
zerland) in DMEM containing 25 m
M
Hepes, pH 7.0 at
37 °C. Inhibitor-containing medium was changed every
6–8 h. After two days of incubation the cells were lyzed and
analyzed by Western blotting.
Construction of expression vectors
Blunt end cDNAs of either bovine furin or rat PC7
were cloned into pIRES1hyg (Clontech), leading to the
expression vectors pIRES1hyg-furin or pIRES1hyg-PC7,
respectively [20].
Cloning of the furin nucleotide sequence
from LoVo cells
Total RNA from LoVo cells was isolated using the RNeasy
kit (Qiagen, Hilden, Germany). The two-step RT-PCR was
carried out using Superscript II (Lifetechnologies) and Taq
DNA polymerase (Promega) with the specific primers
Fur1_for (5¢-GTGGGCCGGAAAGTGAGCCA-3¢)and
Fur2_rev (5¢-CCCTTGTAGGAGATGAGGCC-3¢). The
resulting 1058 bp amplificate was isolated, subcloned in
pUC57 (MBI Fermentas) and sequenced.
Western blot analysis of TACE and ADAM10
Cells were washed and collected with NaCl/P
i
then cells
were suspended in cracking buffer [(5 m
M

AIDA
2.0 (Raytest, Straubenhardt,
Germany).
Isolation and detection of APPsa by Western blot
analysis
Depending on each cell line an appropriate number of cells
was seeded on poly
L
-lysine coated 10-cm dishes and
grown for 20 h close to confluency. Then, cells were
washed twice with serum-free culture medium and incuba-
ted for 4.5 h in serum-free culture medium containing
2m
M
glutamine, 100 UÆmL
)1
penicillin, 100 mgÆmL
)1
streptomycin and 10 lgÆmL
)1
fatty acid-free BSA either
in the absence or presence of 1 l
M
PMA. After collection
of the cell culture supernatant, proteins were precipitated
with a final concentration of 10% trichloroacetic acid by
centrifugation. The pellets were washed twice with ice-cold
acetone, dried and dissolved in 2· Laemmli buffer
containing 100 m
M

furin cDNA. As a control, HEK293 cells were transfected
with the empty expression vector. Cellular membrane
proteins were subjected to Western blot analysis and
immunologically detected proteins corresponding to imma-
ture and mature TACE were densitometrically quantified
(Fig. 1B,C). Whereas the ratio of mature TACE relative to
the immature form in HEK control cells was 131 ± 24%
Fig. 1. Proteolytic processing of TACE by proprotein-convertases. In every case TACE was detected with a rabbit polyclonal antibody.
(A) Inhibition of TACE processing by the inhibitor decanoyl-RVKR-chloromethylketone. The inhibitor was added to a final concentration of
30 l
M
to HEK293 cells. After 48 h the cells were lyzed and membrane proteins were immunoblotted. A representative example of three experiments
is shown. (B) Western blot of endogenous TACE in HEK293 cells stably transfected with vector pIRES1hyg alone (HEK293 control), PC7
(HEK293 + PC7) or furin (HEK293 + furin). Blotted cellular membrane proteins were analyzed. (C) Densitometric analysis of TACE pro-
cessing. The proform of TACE in each cell line was set to 100%. The mature form is expressed as percentage of the proform and as mean ± SD of
three independent experiments. Significance was determined by the t-test (w, P < 0.05). (D) Proteolytic processing of TACE in furin-deficient
LoVo cells. Cells were grown in DMEM nutrient mixture F-12 almost to confluency, then lyzed and membrane proteins were analyzed by Western
blotting.
2388 K. Endres et al.(Eur. J. Biochem. 270) Ó FEBS 2003
the ratio in HEK cells overexpressing either PC7 or furin
was 187 ± 24% and 216 ± 25%, respectively. Thus,
increased amounts of the proteolytically processed mature
form were detected in PC7 as well as in furin overexpressing
cells suggesting that both proprotein-convertases are able to
process TACE. As higher amounts of the mature form were
detected in furin overexpressing cells, it appears that TACE
is a better substrate for furin than for PC7.
On basis of this result we investigated the effect of furin-
deficiency on TACE maturation. As the human carcinoma
cell line LoVo expresses only the proprotein-convertases

within 6 h of PMA treatment (Fig. 2A). This result
implicates a different susceptibility of TACE and ADAM10
turnover to PMA induced signal transduction processes.
Moreover, the amount of mature TACE apparently
decreases linearly with the time of PMA treatment with a
halftime of approximately 6 h (Fig. 2B).
Stimulation of protein kinases C and A do not affect
the amount of mature TACE
As the nonphysiological PKC stimulator PMA decreased
the amount of mature TACE, we were interested in whether
a more physiological pathway for PKC activation causes
similar effects.
HEK293 cells express G protein-coupled muscarinic
receptors and agonist binding results in an intracellular
increase of the second messengers inositol 1,4,5-trisphos-
phate and diacylglycerol. The latter like PMA binds to the
C1b domain of most PKC isoenzymes and activates them.
HEK293 cells were treated with 100 l
M
acetylcholine and
harvested after 4 h as described. However, we did not find
diminished amounts of mature TACE (Fig. 3) although the
used cells responded to the applied ligand with an intracel-
lular calcium-efflux (proved by fura-2/Ca
2+
fluorescence
measurements; data not shown). Therefore, we conclude
Fig. 2. Effect of the phorbol ester PMA on the processing of endogenous
TACE and ADAM10 in HEK293 cells. Cells were incubated for
1.5–6 h in DMEM containing either 1 l

mean ± SD from a characteristic experiment using triplicates. A
representative example of two experiments is shown.
Ó FEBS 2003 TACE maturation and vanishing of its mature form (Eur. J. Biochem. 270) 2389
that the stimulatory effect of acetylcholine was not main-
tained long enough by the cells to induce the long-term
effect on TACE degradation.
As intracellular signaling pathways act as networks and
mutually influence each other we investigated whether the
PKA signaling pathway might be involved in the reduction
of mature TACE. For this purpose we tested the effect of
the cAMP-analogon dibutyryl-cAMP (dB-cAMP), which is
a strong and long-lasting effector of PKA. HEK293 cells
were incubated in medium supplemented with 0.2 m
M
dB-cAMP for 4 h and membrane proteins were analyzed
by immunoblotting. As shown in Fig. 3 dB-cAMP
displayed no effect on the expression and on the amount
of mature TACE. Similar results were obtained for
ADAM10, where also neither dB-cAMP nor acetylcholine
affected its maturation (not shown).
The effect of PMA on mature TACE and ADAM10
in SH-SY5Y and LoVo cells
To demonstrate that the reduction of catalytically active
TACE following phorbol ester stimulation is not restricted
to HEK293, we tested two other cell lines in respect to
APPsa production and TACE as well as ADAM10
maturation: The human SH-SY5Y cell line is of neuronal
origin; the other line LoVo (colon carcinoma) was chosen
because it was described to be insensitive to PMA in the
context of a-secretase activity and APPsa secretion [40].

aberrant termination of the furin polypeptide [37], the other
is a nucleotide exchange, which leads to the amino acid
exchange W547R in the homo B domain of furin [41].
To confirm these mutations, furin mRNA of LoVo cells
was amplified by RT-PCR with suitable primers. The
obtained nucleotide sequence contained the expected nuc-
leotide exchange in the furin mRNA (not shown) confirm-
ing the integrity of the LoVo cell line used in our
experiments. In conclusion, our results indicate that furin
is not necessarily needed for the PMA-induced APP
shedding in LoVo cells.
The lack of furin is not the key for the persistence
of mature TACE in LoVo cells after PMA stimulation
In contrast to the other tested cell lines, a PMA mediated
decrease of mature TACE was not observed in furin-
deficient LoVo cells. To test the possibility that furin
participates in mature TACE degradation, we examined
whether overexpression of functionally active furin in LoVo
cells restores the effect of PMA on the degradation of
mature TACE. Therefore, LoVo cells were reconstituted
with furin by a transient transfection. After 48 h transfected
cells were stimulated with 1 l
M
PMA and cellular mem-
brane proteins were analyzed by immunoblotting. When
compared to mock transfected cells (LoVo Hyg) the amount
of mature TACE was increased in cells that were transfected
with the furin expression vector (LoVo Furin, Fig. 5). This
Fig. 4. Effect of PMA on the processing of endogenous TACE and
ADAM10 and on the APPsa release from HEK293 LoVo and SHY-5Y

The prodomain of the catalytically active members of the
ADAM family is thought to act as an inhibitor of the
proteinase via a cysteine switch mechanism [42,43]. There-
fore removal of the prodomain is required to obtain the
proteolytically active enzyme [14,33,44]. Recently, we have
shown that ADAM10 is proteolytically processed by both
furin and PC7 and that the removal of the prodomain is
accompanied by an enhanced proteolytic activity [20]. For
TACE it has been shown that the maturation occurs during
the transit of the protein through the late Golgi compart-
ment suggesting that prodomain removal is performed by a
furin-type proprotein-convertase [33]. Consistent with this
model, TACE contains a putative proprotein-convertase
cleavage site, which might be used to generate the mature
enzyme [34,35]. Here we demonstrate that proprotein-
convertases are indeed involved in the maturation of TACE
and that pro-TACE is proteolytically processed by both
furin and PC7 to its mature form, most likely to increase its
proteolytical activity. Because higher amounts of mature
TACE could be detected in furin overexpressing cells, it
might be that pro-TACE is a better substrate for furin than
for PC7. However, this observation may also be due to
different expression levels of the proprotein-convertases and
is therefore difficult to substantiate. The examination of
TACE processing in LoVo cells indicates that there is
redundancy in the proteolytic maturation of TACE as other
members of the PC family can compensate a lacking furin
activity. Therefore, we cannot exclude that additional
members of the PC family also contribute to TACE
activation.

messenger inositol 1,4,5-trisphoshate, which is generated
together with diacylglycerol from phosphatidyl inositol
4,5-bisphosphate by PLCb. Obviously, receptor mediated
increase of diacylglycerol and activation of PKC does not
affect the degradation of mature TACE.
An agonist-induced activation of cellular signaling
pathways is a short-term effect. G protein-coupled
receptors are desensitized upon permanent agonist avail-
ability and therefore do not respond any longer to
effector protein activation. As mature TACE degradation
is a long-term effect, the short-term activation of PKC
by diacylglycerol might not cause a similar effect.
Alternatively, our results with acetylcholine stimulation
of cells which had no effect on TACE maturation
indicates that the effect of PMA may be independent of
PKC and may include other PMA binding molecules
such as the Munc proteins, which are involved in vesicle
formation [45].
Intracellular signaling pathways act as networks and are
mutually influenced. Therefore we investigated the effect of
a long-term PKA activation on mature TACE disappear-
ance. Activation of PKA by dibutyryl-cAMP, a more stable
cAMP analogue, did not influence the degradation of
mature TACE indicating that this effect is not dependent
on PKA.
A PMA mediated decrease in the amount of mature
TACE did not occur in the furin-deficient cell line LoVo.
Therefore, we investigated the role of furin in the PMA
mediated decrease of mature TACE.
Furin cycles between the trans-Golgi network and the cell

of TACE in furin transfected LoVo cells compensates to
some extent the effect of a PMA-induced degradation of
mature TACE. As LoVo cells are of carcinoma origin,
another mutation or a chromosomal rearrangement event
could be responsible for the inactivation of the mature
TACE degrading machinery, which is sensitive to phorbol
esters.
Taken together, both TACE and ADAM10 possess
a-secretase activity and are proteolytically activated by PC7
and furin. Furthermore, a furin-independent and PMA
induced disappearance of mature TACE takes place which
is not evident for mature ADAM10.
Thus, mature forms of TACE and ADAM10 differ in
their cellular stability, which may affect their a-secretase
activity in vivo.
Acknowledgements
This work was supported by grants from the Hirnliga e.V.,
the Deutsche Forschungsgemeinschaft (FA-122/4: DFG Priority
Program – Cellular mechanisms of Alzheimer’s disease) and Fonds
der Chemischen Industrie.
References
1. Blobel, C.P., Wolfsberg, T.G., Turck, C.W., Myles, D.G.,
Primakoff, P. & White, J.M. (1992) A potential fusion peptide
andanintegrinliganddomaininaproteinactiveinsperm-egg
fusion. Nature 356, 248–252.
2. Cho, C., O’Dell Bunch, D., Faure, J E., Goulding, E.H., Eddy,
E.M., Primakoff, P. & Myles, D.G. (1998) Fertilization defects in
sperm from mice lacking fertilin b. Science 281, 1857–1859.
3. Yagami-Hiromasa, T., Sato, T., Kurisaki, T., Kamijo, K.,
Nabeshima, Y I. & Fujisawa-Sehara, T. (1995) A metallopro-

3603–3617.
11. Fan, H. & Derynck, R. (1999) Ectodomain shedding of TGF-
alpha and other transmembrane proteins is induced by receptor
tyrosine kinase activation and MAP kinase signaling cascades.
EMBO J. 18, 6962–6972.
12. Prenzel, N., Zwick, E., Daub, H., Leserer, M., Abraham, R.,
Wallasch, C. & Ullrich, A. (1999) EGF receptor transactivation by
G-protein-coupled receptors requires metalloproteinase cleavage
of proHB-EGF. Nature 402, 884–888.
13. Nath, D., Williamson, N.J., Jarvis, R. & Murphy, G. (2001)
Shedding of c-Met is regulated by crosstalk between a G-protein
coupled receptor and the EGF receptor and is mediated by a
TIMP-3 sensitive metalloproteinase. J. Cell Sci. 114, 1213–1220.
14. Loechel, F., Overgaard, M.T., Oxvig, C., Albrechtsen, R. &
Wewer, U.M. (1999) Regulation of human ADAM 12 protease by
the prodomain. J. Biol. Chem. 274, 13427–13433.
15. Milla, M.E., Leesnitzer, M.A., Moss, M.L., Clay, W.C., Carter,
H.L., Miller, A.B., Su, J.L., Lambert, M.H., Willard, D.H.,
Sheeley, D.M., Kost, T.A., Burkhart, W., Moyer, M., Blackburn,
R.K., Pahel, G.L., Mitchell, J.L., Hoffman, C.R. & Becherer, J.D.
(1999) Specific sequence elements are required for the expression
of functional tumor necrosis factor-a-converting enzyme (TACE).
J. Biol. Chem. 274, 30563–30570.
16. Steiner, D.F. (1998) The proprotein-convertases. Curr. Opin.
Chem. Biol. 2, 31–39.
17. Seidah, N.G. & Chretien, M. (1999) Proprotein and prohormone
convertases: a family of subtilases generating diverse bioactive
polypeptides. Brain Res. 848, 45–62.
18. Scha
¨

(1994) Reversal of the Swedish familial Alzheimer’s disease
mutant phenotype in cultured cells treated with phorbol 12,13-
dibutyrate. Neurosci. Lett. 174, 173–176.
26. Buxbaum, J.D., Liu, K N., Luo, Y., Slack, J.L., Stocking, K.L.,
Peschon, J.J., Johnson, R.S., Castner, B.J., Cerretti, D.P. & Black,
R.A. (1998) Evidence that tumor necrosis factor a converting
enzymeisinvolvedinregulateda-secretasecleavage of theAlzheimer
amyloid protein precursor. J. Biol. Chem. 273, 27765–27767.
27. Lammich, S., Kojro, E., Postina, R., Gilbert, S., Pfeiffer, R.,
Jasionowski, M., Haass, C. & Fahrenholz, F. (1999) Constitutive
and regulated a-secretase cleavage of Alzheimer’s amyloid pre-
cursor protein by a disintegrin metalloprotease. Proc.NatlAcad.
Sci. USA 96, 3922–3927.
28. Koike, H., Tomioka, S., Sorimachi, H., Saido, T.C., Maruyama,
K.,Okuyama,A.,Fujisawa-Sehara,A.,Ohno,S.,Suzuki,K.&
Ishiura, S. (1999) Membrane-anchored metalloprotease MDC9
has an alpha-secretase activity responsible for processing the
amyloid precursor protein. Biochem. J. 343, 371–375.
29. Roghani, M., Becherer, J.D., Moss, M.L., Atherton, R.E.,
Erdjument-Bromage, H., Arribas, J., Blackburn, R.K., Weskamp,
G., Tempst, P. & Blobel, C.P. (1999) Metalloprotease-disintegrin
MDC9: intracellular maturation and catalytic activity. J. Biol.
Chem. 274, 3531–3540.
30. Marcinkiewicz, M. & Seidah, N.G. (2000) Coordinated expression
of b-amyloid precursor protein and the putative b-secretase BACE
and a-secretase ADAM10 in mouse and human brain. J. Neuro-
chem. 75, 2133–2143.
31. Merlos-Sua
´
rez, A., Ferna

hemagglutinin by specific peptidyl chloroalkyl ketones. Virology
172, 25–31.
37. Takahashi, S., Kasai, K., Hatsuzawa, K., Kitamura, N., Misumi,
Y., Ikehara, Y., Murakami, K. & Nakayama, K. (1993) A
mutation of furin causes the lack of precursor-processing activity
in human colon carcinoma LoVo cells. Biochem. Biophys. Res.
Commun. 15, 1019–1026.
38. Seidah, N.G., Hamelin, H., Mamarbachi, M., Dong, W., Tadros,
H., Mbikay, M., Chretien, M. & Day, R. (1996) cDNA structure,
tissue distribution, and chromosomal localization of rat PC7, a
novel mammalian proprotein-convertase closest to yeast kexin-
like proteinases. Proc. Natl Acad. Sci. 93, 3388–3393.
39. Doedens, J.R. & Black, R.A. (2000) Stimulation-induced down-
regulation of tumor necrosis factor-a converting enzyme. J. Biol.
Chem. 275, 14598–14607.
40. Lopez-Perez, E., Zhang, Y., Frank, S.J., Creemers, J., Seidah, N.
& Checler, F. (2001) Constitutive a-secretase cleavage of the
b-amyloid precursor protein in the furin-deficient LoVo cell line:
involvement of the pro-hormone convertase 7 and the disintegrin
metalloprotease ADAM10. J. Neurochem. 76, 1532–1539.
41. Takahashi, S., Nakagawa, T., Kasai, K., Banno, T., Duguay, S.J.,
VandeVen,W.J.M.,Murakami,K.&Nakayama,K.(1995)
A second mutant allele of furin in the processing-incompetent cell
line, LoVo. J. Biol. Chem. 270, 26565–26569.
42. van Wart, H.E. & Birkedal-Hansen, H. (1990) The cysteine switch:
a principle of regulation of metalloproteinase activity with
potential applicability to the entire matrix metalloproteinase gene
family. Proc. Natl Acad. Sci. USA 87, 5578–5582.
43.Grams,F.,Huber,R.,Kress,L.F.,Moroder,L.&Bode,W.
(1993) Activation of snake venom metalloproteinases by a cysteine