TNF-α converting enzyme (TACE) is the protease responsible for processing proTNF from the 26-kDa membrane-anchored precursor to the secreted 17-kDa TNF-α. We show here that a deletion mutant of TACE (dTACE), lacking the pro and catalytic domains of the protease, acts as a dominant negative for proTNF processing in transfected HEK293 cells. We used the same system to test the effect of dTACE on TNFRII processing. Overexpression of dTACE with TNFRII resulted in >80% inhibition of TNFRII shedding. Although significant inhibition of TNF-α and TNFRII shedding was achieved with dTACE, we could not detect a cell surface accumulation of the noncleaved substrates above that observed in the absence of dTACE. Our results suggest that TNFRII is a substrate for TACE, and that dTACE is capable of interfering with the function of endogenous TACE, either by binding and sequestering TACE substrates via the disintegrin domain, transmembrane domain, or cytoplasmic tail, or by some other mechanism that has yet to be determined.

Many functionally diverse proteins are synthesized as membrane-anchored precursors that are released from the cell by a regulated process of proteolysis. This shedding process, which often occurs after cell activation (1, 2), can rapidly modulate cell surface responsiveness to external signals as well as release soluble regulatory factors. The shedding of many proteins (2, 3, 4, 5) is blocked by hydroxamic acid-based inhibitors, implying that the enzyme(s) that mediates the release belongs to the zinc-metalloprotease family. It is not yet known whether membrane protein shedding is mediated by several enzyme activities or through a single protein functioning as a general “sheddase.”

One of the most extensively studied shedding events is the release of soluble TNF-α (TNF), a pleiotropic cytokine produced primarily by macrophages and T cells (3, 4, 5). TNF is synthesized as a membrane-anchored 26-kDa precursor (proTNF) that is cleaved to the secreted 17-kDa form. The release of soluble TNF initiates a diverse array of inflammatory and immune modulatory activities necessary for host defense. Clinical interest in TNF has stemmed from its known pathophysiological role in systemic responses (6).

ProTNF cleavage is one of the few shedding events for which a processing enzyme has been identified. TNF-α converting enzyme (TACE)2 is a member of the metalloprotease-disintegrin family of membrane-anchored glycoproteins (7, 8). Its role in proTNF processing has been supported by studies using chimeric mice that are null for TACE activity (7). T cells derived from these mice express a form of TACE that lacks a portion of the catalytic domain and are unable to process proTNF. Recent reports indicate that the shedding of other cell surface proteins is also hindered in this background (9, 10). Because TACE is expressed in these cells, albeit in an enzymatically inactive state, it is unclear whether the effects are due to the absence of TACE activity or to a dominant negative effect on all metalloprotease-mediated cleavages. In this study, we show that a mutant form of TACE lacking the pro and catalytic domains of the protease (referred to as dTACE) acts as a dominant negative for proTNF and TNFRII processing. Possible mechanisms by which all or part of the disintegrin, transmembrane, and cytoplasmic tail regions of dTACE inhibit endogenous TACE function could involve a direct binding and sequestering of TACE substrates or the indirect inhibition of endogenous TACE activation.

Full-length proTNF cDNA was PCR amplified from RNA purified from LPS-treated monocytes and cloned into the mammalian expression vector pcDNA3.1 (Invitrogen, Carlsbad, CA). Recombinant TACE (11) was used as a template to generate a PCR fragment encoding the disintegrin, transmembrane, and cytoplasmic tail regions of enzyme. The 5′ (AAGCTTAGCAATAAAGTTTGTGGGAAC) primer included a HindIII site, and the 3′ (TCTAGATTAGCACTGTGTTTCTTTGC) primer included an XbaI site for cloning into the pFLAG/CMV1 expression vector (Eastman Kodak, Rochester, NY). Full-length TNFRII was PCR amplified from LPS-treated monocytes and subcloned into the HindIII/XbaI site of pcDNA3.

HEK293 cells (American Type Culture Collection, Manassas, VA) were grown in DMEM (Life Technologies, Grand Island, NY) supplemented with 10% FBS (HyClone, Logan, UT), 2 mM glutamine, and penicillin G/streptomycin (Life Technologies, Gaithersburg, MD). Transfections were conducted using lipofectamine reagent (Life Technologies). HEK293 cells (2 × 106) were seeded onto 100-mM tissue culture dishes 24 h before transfection. Cells were washed with OPTI-MEM I reduced serum medium (Life Technologies), incubated with the transfection mixture (10 μg of DNA and 50 μl of lipofectamine reagent) for 6 h at 37°C in a CO2 incubator, and washed once with PBS before being replaced with RPMI 1640 growth medium. Cells were washed at 24 h posttransfection before fresh medium was added. Soluble TNF and TNFRII were allowed to accumulate in the culture medium for 24 or 8 h, respectively, before culture supernatant harvest. For transfections with proTNF or TNFRII cDNA, a total of 1 μg of each of these DNAs was added to 9 μg of pcDNA3.1 carrier DNA. Cotransfection of proTNF or TNFRII with dTACE cDNA was conducted using 9 μg of dTACE cDNA with 1 μg of either proTNF or TNFRII cDNA.

The level of secreted TNF produced from transfected HEK293 cells was quantitated by ELISA (12). Soluble TNFRII was quantitated using the human soluble TNFRII Quantikine immunoassay from R&D Systems (Minneapolis, MN).

ProTNF-transfected HEK293 cells were lifted from plates at 48 h posttransfection (1× EDTA/trypsin (Life Technologies)), washed with PBS, and resuspended at 5 × 106 cells/ml. Cells (5 × 105) were incubated with either 10 μg/ml of TNF-specific mAb or an isotype-matched mouse IgG (mIgG) for 30 min at 4°C, washed with PBS, and treated with a PE-conjugated goat anti-mouse IgG (Chemicon, Temecula, CA). Cells were washed once before resuspension in PBS and analysis on a Becton Dickinson FACScan (Mountain View, CA). The Abs MAB266 (R&D) and M2 (Sigma, Costa Mesa, CA) were used to characterize the expression of TNFRII and dTACE, respectively.

Metabolic labeling and immunoprecipitation experiments were conducted as described previously (12). HEK293 cells, transfected with either proTNF alone or in combination with dTACE, were harvested (2 × 106 each) at 48 h posttransfection and pulse-labeled with [35S]cysteine for 30 min before lysis. Immunoprecipitations were conducted using the TNF-specific polyclonal Ab Rb504. Immunoprecipitates were subjected to electrophoresis in a 14% SDS-polyacrylamide gel and visualized by autoradiography.

In this study, we sought to further investigate whether TACE, in addition to its known role in proTNF processing, could mediate the shedding of other membrane proteins. For these studies we chose TNFRII as a candidate for TACE processing, because the cell surface release of TNFRII and TNF occur concurrently, are similarly sensitive to inhibition by the hydroxamate TAPI (5), and possess a similar cleavage site (alanine-valine). Our approach was to express a mutant form of TACE that we hypothesized would act as a dominant negative for endogenous TACE activity, previously shown to exist in HEK293 cells (13). We designed our TACE mutant construct after an N-terminal truncation of the homologous Kuzbanian protein (14), which encoded only the disintegrin, transmembrane, and cytoplasmic tail regions of the enzyme. This proteolytically inactive mutant acted as a dominant negative form for Notch processing in Drosophila. PCR amplification of a comparable fragment of TACE was performed using recombinant full-length TACE as a template, and the sequence structure proposed by Black et al. (7) as a guide for domain delineation. The resultant product (encoding amino acids S 474 to C 824) was cloned into the pFLAG/CMV1 vector, which drives the expression of a TACE fusion protein containing an N-terminal preprotrypsin signal sequence followed by the 8-aa FLAG tag. We refer to this mutant form of TACE as dTACE.

Before testing the functional activity of dTACE, we confirmed endogenous TACE activity in HEK293 cells using criteria described previously (13). Furthermore, because recent reports implicate ADAM 10 as a TNF convertase (15, 16), we analyzed the HEK293 cells for ADAM 10 expression by RT-PCR. HEK293 cells express ADAM 10 at levels comparable with that expressed in human peripheral blood monocytes (data not shown). Therefore, HEK293 cells express at least two convertases reported to be capable of cleaving proTNF.

Given this information, we used the HEK293 background to test the effect of dTACE expression on proTNF processing. Recombinant proTNF was transfected into HEK293 cells with or without dTACE cDNA. TNF secretion into the culture supernatant was quantified after an accumulation period of 24 h. HEK293 cells, on average, produced 18.5 ng/ml (range 11.5–25 ng/ml/4 × 106 cells) of soluble TNF when transfected with 1 μg of proTNF cDNA (Fig. 1). This level of TNF secretion was reduced by 84% in cells overexpressing the mutant dTACE, which was comparable with that obtained with hydroxamic acid-based metalloprotease inhibitor treatment. Addition of PMA, which increased the soluble TNF released from TNF transfected cells by 128%, did not overcome the inhibitory effects of dTACE on the metalloprotease-mediated cleavage of proTNF (data not shown). The inhibition of proTNF secretion occurred in a dTACE dose-dependent manner, because decreasing the amount of dTACE cDNA from 9 μg to 0.5 μg reduced the inhibitory effect from 90% to 30%, respectively (Fig. 2). We confirmed the expression of dTACE and proTNF in each of the transfection experiments by monitoring the cell surface levels of each protein by immunofluorescence labeling (Fig. 3).

FIGURE 1.

Effect of dTACE on TNF secretion. HEK293 cells were transfected with proTNF cDNA in combination with either pcDNA3.1 (proTNF alone) or dTACE cDNA (proTNF + dTACE). ProTNF + HA represents HEK293 cells transfected with proTNF (plus carrier) and treated with a broad-based hydroxamic acid for 24 h before supernatant harvest. Data are a compilation of nine separate experiments.

FIGURE 1.

Effect of dTACE on TNF secretion. HEK293 cells were transfected with proTNF cDNA in combination with either pcDNA3.1 (proTNF alone) or dTACE cDNA (proTNF + dTACE). ProTNF + HA represents HEK293 cells transfected with proTNF (plus carrier) and treated with a broad-based hydroxamic acid for 24 h before supernatant harvest. Data are a compilation of nine separate experiments.

Close modal
FIGURE 2.

Dose-dependent inhibition of TNF secretion by dTACE. The percentage of inhibition of TNF-α production from HEK293 cells transfected with proTNF (1 μg) and increasing amounts of dTACE cDNA (0, 0.5, 1, 2, 4, 5, and 9 μg) is shown.

FIGURE 2.

Dose-dependent inhibition of TNF secretion by dTACE. The percentage of inhibition of TNF-α production from HEK293 cells transfected with proTNF (1 μg) and increasing amounts of dTACE cDNA (0, 0.5, 1, 2, 4, 5, and 9 μg) is shown.

Close modal
FIGURE 3.

Cell surface expression of proTNF and dTACE in cotransfected cells. Cells were incubated either with mIgG or anti-TNF-α mAb (A) or with mIgG or anti-FLAG (dTACE) mAb (B) for 30 min before incubation with anti-mouse PE Ab and visualization by flow cytometry.

FIGURE 3.

Cell surface expression of proTNF and dTACE in cotransfected cells. Cells were incubated either with mIgG or anti-TNF-α mAb (A) or with mIgG or anti-FLAG (dTACE) mAb (B) for 30 min before incubation with anti-mouse PE Ab and visualization by flow cytometry.

Close modal

To ensure that the inhibitory effects of dTACE on TNF secretion were due to the presence of the mutant form of TACE and not to reduced proTNF substrate in cells expressing two recombinant proteins, we metabolically labeled proTNF and proTNF/dTACE transfected cells before lysis and immunoprecipitation of proTNF protein. The precipitates were analyzed on a SDS gel and visualized by autoradiography (Fig. 4). The results indicate that cells transfected with proTNF alone or in combination with dTACE express comparable levels of proTNF protein. We conclude that the reduced level of TNF secretion from cells expressing dTACE is not due to reduced proTNF protein synthesis but rather is due to a true inhibition of proTNF processing.

FIGURE 4.

Comparison of TNF expression by proTNF and proTNF + dTACE transfectants. ProTNF (1) or proTNF + dTACE (2) transfected cells were pulsed with [35S]cysteine before lysis, immunoprecipitation, and analysis by SDS-PAGE and autoradiography.

FIGURE 4.

Comparison of TNF expression by proTNF and proTNF + dTACE transfectants. ProTNF (1) or proTNF + dTACE (2) transfected cells were pulsed with [35S]cysteine before lysis, immunoprecipitation, and analysis by SDS-PAGE and autoradiography.

Close modal

Thus, an N-terminal truncation of TACE, comprising the disintegrin, transmembrane, and cytoplasmic tail region of the enzyme, is capable of preventing the processing of proTNF by endogenous TACE and, presumably, ADAM 10 in HEK293 cells. Therefore, we classify dTACE as dominant negative for metalloprotease-mediated cleavage of proTNF in this cellular background.

We subsequently tested the effect of dTACE on the release of soluble TNFRII from transfected HEK293 cells. HEK293 cells constitutively secrete TNFRII at an average (n = 2) level of 1.43 ng/ml/4 × 106 cells over the 8-h accumulation period when transfected with recombinant TNFRII alone (Fig. 5). Coexpression of dTACE with TNFRII resulted in an 88% reduction in the level of soluble TNFRII released (Fig. 5), which was comparable with that obtained in the presence of PMA (data not shown). The cell surface expression of TNFRII and dTACE was confirmed in each of the transfections by immunofluorescent staining followed by FACScan analysis (Fig. 6). These data indicate that dTACE can inhibit the proteolytic release of both proTNF and TNFRII.

FIGURE 5.

Effect of dTACE on TNFRII secretion. HEK293 cells were transfected with TNFRII cDNA alone or with dTACE cDNA. Soluble TNFRII was allowed to accumulate for 8 h before quantification (ELISA). Data represent a compilation of three separate experiments.

FIGURE 5.

Effect of dTACE on TNFRII secretion. HEK293 cells were transfected with TNFRII cDNA alone or with dTACE cDNA. Soluble TNFRII was allowed to accumulate for 8 h before quantification (ELISA). Data represent a compilation of three separate experiments.

Close modal
FIGURE 6.

Cell surface expression of TNFRII and dTACE in cotransfected cells. Cells incubated either with mIgG or anti-TNFRII mAb (A) or with mIgG and anti-FLAG (dTACE) mAb (B) before incubation with anti-mouse PE Ab and visualization by flow cytometry are shown.

FIGURE 6.

Cell surface expression of TNFRII and dTACE in cotransfected cells. Cells incubated either with mIgG or anti-TNFRII mAb (A) or with mIgG and anti-FLAG (dTACE) mAb (B) before incubation with anti-mouse PE Ab and visualization by flow cytometry are shown.

Close modal

Because one of the consequences of inhibiting the shedding of membrane-bound proteins is the potential for accumulation on the cell surface, we used flow cytometry to probe the effect of dTACE on proTNF and TNFRII cell surface expression (Table I). Transfection of HEK293 cells with either proTNF or TNFRII cDNA resulted in a significant increase in the geometric mean fluorescence values over background. Interestingly, we could not detect a measurable increase in the cell surface expression of proTNF or TNFRII with coexpression of dTACE under conditions that inhibited release into the media by ∼90%. Therefore, inhibition of TNFRII and proTNF shedding by dTACE does not induce an accumulation of cell surface expression of either of these proteins over that observed in cells lacking the dominant negative form of TACE. In a previous report, we provided evidence for rapid degradation of unprocessed proTNF in LPS-stimulated monocytes treated with hydroxamic-based metalloprotease inhibitors (12). It appears that the proTNF and TNFRII transfected HEK293 cells may invoke a similar mechanism of degradation of noncleaved substrates.

Table I.

Effect of dTACE on proTNF and TNFRII cell surface expressiona

TreatmentmIgGProTNFTNFRIIdTACE
ProTNF alone 14 32 — 15 
ProTNF+ dTACE 14 46 — 121 
TNFRII alone 16 — 134 15 
TNFRII+ dTACE 15 — 120 137 
TreatmentmIgGProTNFTNFRIIdTACE
ProTNF alone 14 32 — 15 
ProTNF+ dTACE 14 46 — 121 
TNFRII alone 16 — 134 15 
TNFRII+ dTACE 15 — 120 137 
a

Transfected cells were monitored for cell surface expression of proTNF, TNFRII, and dTACE by flow cytometry. Data are represented as mean fluorescence intensity values.

In summary, these results indicate that an N-terminally truncated and enzymatically inactive form of TACE is capable of inhibiting metalloprotease-mediated cleavage of proTNF and TNFRII in a cellular background that expresses endogenous TACE as well as other family members shown to be capable of processing proTNF (i.e., ADAM 10). The dominant negative effects we observed are likely due to the interaction of all or part of the disintegrin, transmembrane, and cytoplasmic tail of TACE with proTNF and TNFRII, thereby preventing association and cleavage by the endogenous metalloprotease activity. However, because a direct association of TACE with the two substrates remains to be established, alternative explanations for the observed effects are possible, including an inhibition of TACE activation by cytoplasmic proteins such as protein kinases or the sequestering of cofactors required for enzyme activity. The mechanism of inhibition, as well as the minimal portion of TACE required for these effects, will need to be probed with further experimentation. Finally, because we know that an enzymatically inactive form of TACE can act as a dominant negative, it is possible that the global inhibition of shedding events observed in cells derived from chimeric mice described by Black and colleagues. (7, 9, 10) may be due to similar dominant negative effects. The role of TACE in cell surface protein shedding must therefore await the generation of mice proven to be completely deficient in TACE expression.

We thank Sherrill Nurnberg, Persymphonie Miller, and Xiaosui Jiang for excellent technical assistance.

1

Address correspondence and reprint requests to Dr. Kimberly A. Solomon, Department of Inflammatory Diseases Research, E400-4241, DuPont Pharmaceuticals Company, P.O. Box 80400, Wilmington, DE 19880-0400.

2

Abbreviations used in this paper: TACE, TNF-α converting enzyme; dTACE, deletion mutant of TACE; TNFRII, TNF receptor II; mIg, mouse Ig; ADAM, a disintegrin and metalloprotease; TAPI, TNF-α protease inhibitor (N-{dl-[2-(hydroxyaminocarbonyl)methyl]-4-methylpentanoyl}-l-3-(2[prime]-naphthyl)alanyl-l-alanine, 2-aminoethylamide); FLAG, marker octapeptide (N-Asp-Tyr-Lys-Asp-Asp-Asp-Asp-Lys-C).

1
Hooper, N. M., E. H. Karran, A. J. Turner.
1997
. Membrane protein secretases.
Biochem. J.
321
:
265
2
Heaney, M. L., D. W. Golde.
1998
. Soluble receptors in human disease.
J. Leukocyte Biol.
64
:
135
3
Gearing, A., P. Beckett, M. Christodoulou, M. Churchill, J. Clements, A. Davidson, A. Drummond, W. Galloway, R. Gilbert, J. Gordon, et al
1994
. Processing of tumour necrosis factor-α precursor by metalloproteinases.
Nature
370
:
555
4
McGeehan, G., J. Becherer, R. Bast, Jr, C. Boyer, B. Champion, K. Connolly, J. Conway, P. Furdon, S. Karp, S. Kidao, et al
1994
. Regulation of tumour necrosis factor-α processing by a metalloproteinase inhibitor.
Nature
370
:
558
5
Crowe, P. D., B. N. Walter, K. M. Mohler, C. Otten-Evans, R. A. Black, C. F. Ware.
1995
. A metalloprotease inhibitor blocks shedding of the 80-kDa TNF receptor and TNF processing in T lymphocytes.
J. Exp. Med.
181
:
1205
6
Eigler, A., B. Sinha, G. Hartmann, S. Endres.
1997
. Taming TNF: strategies to restrain this proinflammatory cytokine.
Immunol. Today
18
:
487
7
Black, R. A., C. T. Rauch, C. J. Kozlosky, J. J. Peschon, J. L. Slack, M. F. Wolfson, B. J. Castner, K. L. Stocking, P. Reddy, S. Srinivasan, et al
1997
. A metalloproteinase disintegrin that releases tumour-necrosis factor-α from cells.
Nature
385
:
729
8
Moss, M., S. Jin, M. Milla, W. Burkhart, H. Carter, W. Chen, W. Clay, J. Didsbury, D. Hassler, C. Hoffman, et al
1997
. Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-α.
Nature
385
:
733
9
Buxbaum, J. D., K. N. Liu, Y. Luo, J. L. Slack, K. L. Stocking, J. J. Peschon, R. S. Johnson, B. J. Castner, D. P. Cerretti, R. A. Black.
1998
. Evidence that tumor necrosis factor α converting enzyme is involved in regulated α-secretase cleavage of the Alzheimer amyloid protein precursor.
J. Biol. Chem.
273
:
27765
10
Peschon, J. J., J. L. Slack, P. Reddy, K. L. Stocking, S. W. Sunnarborg, D. C. Lee, W. E. Russell, B. J. Castner, R. S. Johnson, J. N. Fitzner, et al
1998
. An essential role for ectodomain shedding in mammalian development.
Science
282
:
1281
11
Patel, I. R., M. G. Attur, R. N. Patel, S. A. Stuchin, R. A. Abagyan, S. B. Abramson, A. R. Amin.
1998
. TNF-α convertase enzyme from human arthritis-affected cartilage: isolation of cDNA by differential display, expression of the active enzyme, and regulation of TNF-α.
J. Immunol.
160
:
4570
12
Solomon, K. A., M. B. Covington, C. P. DeCicco, R. C. Newton.
1997
. The fate of pro-TNF-α following inhibition of metalloprotease-dependent processing to soluble TNF-α in human monocytes.
J. Immunol.
159
:
4524
13
Glaser, K. B., M. Falduto, R. Metzger, T. Pederson, L. Pease, K. Shiosaki, D. W. Morgan.
1997
. Expression, release, and regulation of human TNF α from stable transfectants of HEK-293 cells.
Inflamm. Res.
46
: (Suppl. 2):
S127
14
Pan, D., G. M. Rubin.
1997
. Kuzbanian controls proteolytic processing of Notch and mediates lateral inhibition during Drosophila and vertebrate neurogenesis.
Cell
90
:
271
15
Lunn, C. A., X. Fan, B. Dalie, K. Miller, P. J. Zavodny, S. K. Narula, D. Lundell.
1997
. Purification of ADAM 10 from bovine spleen as a TNFα convertase.
FEBS Lett.
400
:
333
16
Rosendahl, M., S. Ko, D. Long, M. Brewer, B. Rosenzweig, E. Hedl, L. Anderson, S. Pyle, J. Moreland, M. Meyers, et al
1997
. Identification and characterization of a pro-tumor necrosis factor-α-processing enzyme from the ADAM family of zinc metalloproteases.
J. Biol. Chem.
272
:
24588