SEARCH     advanced search my account e-alerts subscribe register
Journal home
Advance online publication
Current issue
Press releases
For authors
For referees
Contact editorial office
About the journal
For librarians
Contact NPG
Customer services
Site features
NPG Subject areas
Access material from all our publications in your subject area:
Biotechnology Biotechnology
Cancer Cancer
Chemistry Chemistry
Dentistry Dentistry
Development Development
Drug Discovery Drug Discovery
Earth Sciences Earth Sciences
Evolution & Ecology Evolution & Ecology
Genetics Genetics
Immunology Immunology
Materials Materials Science
Medical Research Medical Research
Microbiology Microbiology
Molecular Cell Biology Molecular Cell Biology
Neuroscience Neuroscience
Pharmacology Pharmacology
Physics Physics
Browse all publications
20 June 2002, Volume 21, Number 27, Pages 4237-4246
Table of contents    Previous  Article  Next   [PDF]
Original Paper
Multiple tristetraprolin sequence domains required to induce apoptosis and modulate responses to TNFalpha through distinct pathways
Barbra A Johnson and T Keith Blackwell

Center for Blood Research and Department of Pathology, Harvard Medical School, 200 Longwood Avenue, Boston, Massachusetts, MA 02115, USA

Correspondence to: T K Blackwell, Center for Blood Research and Department of Pathology, Harvard Medical School, 200 Longwood Avenue, Boston, Massachusetts, MA 02115, USA. E-mail: blackwell@cbr.med.harvard.edu


Expression of the immediate early protein tristetraprolin (TTP) is induced by numerous stimuli, including tumor necrosis factor-alpha (TNFalpha). Evidence indicates that TTP limits production of TNFalpha and other cytokines by directly binding and destabilizing their mRNAs. This effect seems to require only the conserved TTP zinc finger region, and is characteristic of the related proteins TIS11b and TIS11d. TTP, TIS11b, and TIS11d each also induce apoptosis through the mitochondrial pathway analogously to certain oncogenes, suggesting that they influence growth or survival signals. Among TTP/TIS11 proteins, TTP alone also promotes apoptosis synergistically with TNFalpha. Here we show that other regions of TTP along with the zinc fingers are required for TTP to induce apoptosis. We also demonstrate that TTP acts through an additional pathway to sensitize cells to the pro-apoptotic stimulus of TNFalpha. This modulation of TNFalpha responses specifically requires the TTP N-terminal region, which is not conserved in TIS11b or TIS11d. We conclude that the physiological functions of TTP depend upon multiple regions of the TTP protein, that TTP has diverged functionally from TIS11b and TIS11d, and that modulation of TNFalpha responses may be a unique and important aspect of TTP function.

Oncogene (2002) 21, 4237-4246 doi:10.1038/sj.onc.1205526


tristetraprolin; TNFalpha; immediate-early; apoptosis; zinc finger


Tristetraprolin (TTP; also Nup475 and TIS11) is an immediate-early protein that is expressed transiently in response to numerous extracellular agents, including the cytokine TNFalpha (Carballo et al., 1998; DuBois et al., 1995; Lai et al., 1990; Varnum et al., 1989). In mice, disruption of the TTP gene causes a general inflammatory and arthritic syndrome which is mediated largely by TNFalpha, and is associated with increased levels and half-life of the TNFalpha mRNA (Carballo et al., 1997; Taylor et al., 1996a). The localization, stability, and translation of TNFalpha and other cytokine mRNAs are regulated through an AU-rich element (ARE) that is located within their respective 3' untranslated regions, and responds to multiple signaling pathways (Chen and Shyu, 1995; Dumitru et al., 2000; Kontoyiannis et al., 1999, 2001; Kotlyarov et al., 1999; Rutault et al., 2001; Vasudevan and Peltz, 2001; Winzen et al., 1999). TTP binds directly to the AREs of TNFalpha and other cytokine mRNAs in vitro, and in transfection assays TTP acts through the ARE to destabilize these mRNAs (Carballo et al., 1998, 2000; Chen et al., 2001; Lai et al., 1999; Mahtani et al., 2001; Ming et al., 2001; Raghavan et al., 2001). These findings suggest that TTP modulates cytokine production through direct ARE-dependent mRNA destabilization (Carballo et al., 1998, 2000; Lai et al., 1999).

TTP and the related proteins TIS11b and TIS11d (TTP/TIS11 proteins) consist of two conserved tandem Cys-X8-Cys-X5-Cys-X3-His (CCCH) zinc fingers, along with divergent N- and C-terminal regions (Figure 1a,c). Other CCCH zinc finger proteins have been implicated in control of mRNA transcription, processing, or translation (Bai and Tolias, 1996; Barabino et al., 1997; Batchelder et al., 1999; Guedes and Priess, 1997; Tabara et al., 1999; Tenenhaus et al., 2001; Zuo and Maniatis, 1996). In transfection assays, the TNFalpha mRNA is bound, deadenylated, and apparently destabilized by all three TTP/TIS11 proteins and, surprisingly, by a 77 amino acid TTP fragment that contains the TTP zinc finger region (Lai et al., 2000). These last findings raise the question of whether other regions of the TTP protein contribute to essential aspects of its physiological functions.

Other evidence suggests that TTP has additional functions besides influencing cytokine mRNA stability. A wide variety of growth factors and mitogens induce transient TTP expression in numerous cell types, and in mice TTP is expressed in regenerating small intestine, liver, and lung, as well as in hematopoietic tissues (Carballo et al., 1998; DuBois et al., 1995, 1990; Ehrenfried et al., 1995; Heximer and Forsdyke, 1993; Landesberg et al., 2001; Nakajima and Wall, 1991; Taylor et al., 1996a; Varnum et al., 1991). TTP expression is also induced during responses to some apoptotic stimuli, including BRCA1 expression and withdrawal of growth factors from neuronal cells (Haas et al., 1993; Harkin et al., 1999; Mesner et al., 1995; Mittelstadt and DeFranco, 1993). We have determined that continuous expression of each TTP/TIS11 protein induces various primary and transformed cell types to undergo apoptosis (Johnson et al., 2000). This response depends upon the mitochondrial apoptosis pathway and is enhanced by growth factor deprivation, analogously to cell death induced by c-Myc, E1A, and certain other oncogenes. These observations together suggest that TTP influences growth or survival pathways, and that TTP, like some other immediate early genes, may have a physiological role in apoptosis induction (Johnson et al., 2000).

When TTP is expressed at levels that are insufficient to cause widespread cell death, it induces apoptosis synergistically with TNFalpha (Johnson et al., 2000). Synergistic induction of apoptosis in combination with TNFalpha is also a property of the immediate-early oncoprotein c-Myc, the pro-apoptotic factor E2F-1, and the adenovirus oncogene E1A (Chen et al., 1987; Janicke et al., 1994; Klefstrom et al., 1997; Phillips et al., 1999). In striking contrast to TTP however, neither TIS11b nor TIS11d act synergistically with TNFalpha (Johnson et al., 2000). This is the only functional difference that has been identified among TTP/TIS11 proteins, suggesting that this synergy between TTP and TNFalpha may reflect a physiological function carried out by TTP alone. TNFalpha causes apoptosis by binding to its Type I receptor and activating caspase 8, an apoptotic stimulus that under certain circumstances can be amplified by activation of the mitochondrial apoptosis pathway (Green, 2000). Simultaneously, TNFalpha stimulates pathways that promote cell survival, proliferation, and activation (Baker and Reddy, 1996; Green, 2000). To understand how TTP and TNFalpha act synergistically, it is essential to determine whether TTP and TNFalpha simply cooperate to activate the mitochondrial apoptosis pathway, or whether TTP influences the balance between the survival and apoptotic stimuli of TNFalpha.

In this study, we have investigated how different regions of the TTP protein contribute to apoptosis induced solely by TTP expression, and by TTP acting in conjunction with TNFalpha. We find that the TTP zinc finger region fails to induce apoptosis independently of either the N- or C-terminal regions, suggesting that multiple regions of the TTP protein are critical for its functions in vivo. We have also determined that TTP and TNFalpha induce apoptosis synergistically independently of mitochondrial involvement, indicating that TTP induces apoptosis on its own and modulates responses to TNFalpha through distinct pathways. Finally, both the zinc finger and unique N-terminal regions of TTP are required to sensitize cells to the apoptotic stimulus of TNFalpha, suggesting that this activity of TTP reflects a specific functional divergence between TTP and other TTP/TIS11 proteins.


Multiple regions of TTP are required for its induction of apoptosis

The only predicted functional domains within TTP are its tandem zinc fingers, but outside of these elements TTP is highly conserved between mouse and human (Figure 1a). This similarity suggests that other regions of TTP in addition to the zinc fingers are likely to be important for TTP functions in vivo, and may be required for TTP to induce apoptosis. To test this model, we constructed a series of TTP truncation mutants by dividing the protein into N-terminal, zinc finger and C-terminal regions, and retaining overlapping sequences to minimize possible disruption of functional elements (Figure 1b). For example, the 98 amino acid TTP(Zn) mutant contained flanking residues on both sides of the two CCCH zinc fingers. We then assayed these TTP mutants for induction of apoptosis by transfecting them into 3T3 cells as described previously (Johnson et al., 2000).

When expressed continuously by transfection, TTP induces cell death that is associated with TUNEL positivity and pyknotic nuclei, and is inhibited by Bcl-2 co-expression, indicating apoptosis that is dependent upon the mitochondrial pathway (Figure 2a; not shown) (Johnson et al., 2000). The levels at which TTP is expressed under these conditions overlap with levels to which endogenous TTP is induced transiently in response to extracellular stimuli (Johnson et al., 2000). In addition, significant apoptosis is associated with low level expression of TTP, suggesting that this cell death derives from continuous presence of TTP and does not require excessive TTP expression (Johnson et al., 2000). Induction of apoptosis by TTP was nearly abrogated by point mutation of both zinc fingers (M1,2; Figure 2b), as noted previously (Johnson et al., 2000), and was reduced but not eliminated by deletion of either the N- or C- terminus (TTP(N-Zn) and TTP(Zn-C), respectively, Figure 2b). Like TTP, both TTP(N-Zn) and TTP(Zn-C) triggered apoptosis that could be inhibited by Bcl-2, suggesting dependence on the mitochondrial pathway (Figure 2a). As observed previously for TTP (Johnson et al., 2000), in each case cell death was also attenuated by the caspase inhibitor CrmA, suggesting that death receptors may be involved. In contrast, the TTP zinc fingers on their own were incapable of inducing apoptosis (TTP(Zn); Figure 2b), even when a 10-fold greater amount of expression vector was introduced (not shown).

We verified expression of these TTP mutants by Western blotting, and by similarly analysing fusion proteins in which TTP mutants were linked at their N-terminus to green fluorescent protein (GFP). Analysis with an antibody against the TTP C-terminus (cTTP) indicated that TTP, TTP(C), TTP(Zn-C), and M1,2 were expressed at comparable levels (Figure 2c, lanes 1 and 4-6). Multiple lines of evidence indicated that the TTP(Zn) and TTP(N-Zn) mutants were also expressed comparably but were not effectively detected by cTTP, presumably because the cTTP epitope is not efficiently recognized when it is present ectopically near the zinc finger region (Figure 2c, lanes 2 and 3). Parallel analysis with an antibody against the TTP N-terminus (nTTP) revealed that TTP(N-Zn) was actually produced at somewhat higher levels than TTP (Figure 2c, lanes 7 and 8). By both Western blotting and immunofluorescence, cTTP detected TTP(Zn) but not TTP(N-Zn) (Figure 2c, lanes 2 and 3; Figure 3b; not shown), indicating that TTP(Zn) was expressed at higher levels. Finally, analysis with a GFP antibody revealed that in this transfection assay, GFP-TTP(Zn) was expressed more robustly than either GFP-TTP or GFP-TTP(Zn-C) (Figure 2e, lanes 1-3; bottom panel). Like their non-tagged counterparts, GFP-TTP(Zn-C) and GFP-TTP each induced significant apoptosis, and GFP-TTP(Zn) failed to induce significant cell death (Figure 2d). Together, our findings demonstrate that the N- and C-terminal regions of TTP, in addition to its zinc fingers, are each important for induction of apoptosis.

The failure of the TTP(Zn) truncation mutant to induce apoptosis could derive from its being mis-localized to a different cellular compartment from TTP mutants that are active in this assay. To address this possibility, we used immunofluorescence to investigate how TTP mutants are localized within transfected 3T3 cells (Figure 3a,b). It has been reported that in hematopoietic cell lines TTP is present primarily in the cytoplasm, and that in quiescent fibroblasts TTP is largely nuclear, but is localized to the cytoplasm upon treatment with growth factors (Taylor et al., 1995, 1996b). Consistent with these observations, in nearly all transfected cycling 3T3 cells TTP staining was either predominantly cytoplasmic, or was distributed in a generalized fashion (Figure 3a,b). Mutation of the zinc fingers decreased the fraction of cells in which TTP was localized solely within the cytoplasm (M1,2, Figure 3a,b), indicating that the functional integrity of the zinc fingers is important for the normal distribution of TTP within the cell. TTP(N-Zn), TTP(Zn-C), and TTP(Zn) were each localized predominantly to the nucleus (64, 97, and 96% nuclear respectively; Figure 3a,b), but while TTP(N-Zn) and TTP(Zn-C) induced apoptosis, TTP(Zn) did not (Figure 2b,d). The finding that TTP(Zn) and TTP(Zn-C) differed dramatically in induction of apoptosis but were localized similarly (Figures 2a,d; 3a,b) suggests that the inactivity of TTP(Zn) in this assay does not derive simply from its localization within the cell.

TTP sensitizes cells to TNFalpha-induced apoptosis through a different mechanism

Among the three TTP/TIS11 proteins, a single functional difference has been identified: only TTP induces apoptosis synergistically with TNFalpha (Johnson et al., 2000). Because this functional synergy is a specific property of TTP, it would be predicted to require sequences that vary among TTP/TIS11 proteins. In vertebrates, TTP, TIS11b, and TIS11d can be distinguished from each other by amino acids within their respective zinc finger regions (Lai et al., 2000) (Figure 1c), but these proteins are far more divergent within regions that flank the zinc fingers (Figure 1a). We have used the TTP deletion mutants shown in Figure 1b to determine whether these divergent flanking regions are required for TTP to act synergistically with TNFalpha.

As reported previously, when TTP was expressed at levels that cause only modest apoptosis, treatment with low concentrations of TNFalpha resulted in a significant increase in cell death (Figure 4a). In contrast, the TTP(Zn) and M1,2 mutants (Figure 1b) each failed to synergize with TNFalpha to induce apoptosis (Figure 4a). No significant TNFalpha-dependent apoptosis occurred when TTP(Zn) was expressed at much higher levels (Figure 4c,d), further supporting the idea that the TTP zinc fingers are necessary but not sufficient for this effect. TTP(Zn-C) similarly failed to promote TNFalpha-dependent apoptosis (Figure 4a), even when it was expressed at levels that caused significant apoptosis in the absence of TNFalpha (Figure 4c,d). In striking contrast, TTP(N-Zn) was reproducibly comparable to TTP in promoting TNFalpha-dependent apoptosis (Figure 4a,c,d; not shown). When activation of survival factors by TNFalpha was suppressed by cycloheximide, low concentrations of TNFalpha induced rapid apoptosis in TTP-expressing cells (Figure 4b). In the presence of cycloheximide, TNFalpha treatment similarly caused rapid apoptosis of cells that expressed TTP(N-Zn), but TTP(Zn-C) was only marginally active (Figure 4b). We conclude that the TTP N-terminal and zinc finger regions are together specifically required for TTP to promote apoptosis synergistically with TNFalpha.

TTP and TNFalpha could induce apoptosis synergistically by acting together through the mitochondrial apoptosis pathway, which is required for TTP/TIS11 proteins to induce cell death independently of TNFalpha (Johnson et al., 2000). Alternatively, and of greater potential significance, TTP could influence the apoptotic or survival signaling pathways that are activated directly by TNFalpha. Both models predict that apoptosis derived from the synergistic effects of TTP and TNFalpha should be attenuated by CrmA, which inhibits cell death that occurs in response to either TNFalpha or TTP (Figure 2a) (Nakajima and Wall, 1991; Ray et al., 1992). Accordingly, CrmA suppressed TNFalpha-dependent apoptosis in cells that expressed either TTP or TTP(N-Zn) (Figure 5a,b). If the synergistic effects of TTP and TNFalpha are mediated through the mitochondrial apoptosis pathway, the resulting synergistic increase in cell death should also be inhibited by Bcl-2, which robustly suppresses apoptosis caused by either TTP or TTP(N-Zn) alone (Figure 2a). In striking contrast, Bcl-2 expression failed to inhibit cell death caused by the synergistic effects of TTP or TTP(N-Zn) together with TNFalpha (Figure 5a,b). This last finding suggests that the mitochondrial pathway is not required for TTP and TNFalpha to induce apoptosis synergistically, implying that this cell death is caused by direct caspase activation by TNFalpha. We conclude that TTP sensitizes cells to this apoptotic stimulus, presumably by affecting signaling pathways directly downstream of TNFalpha (Figure 6).


Multiple TTP domains are required for its functions

Dose-response transfection studies suggest that the TTP zinc finger region can independently bind and destabilize the TNFalpha mRNA (Lai et al., 2000). This indicates that the remainder of the TTP protein might not be essential for its functional activity, but may instead play regulatory roles. In apparent contrast, we have observed that the isolated TTP zinc finger region fails to induce apoptosis or to sensitize cells to the apoptotic stimulus of TNFalpha (Figures 2 and 4). Multiple lines of evidence indicate that the inactivity of the TTP zinc finger region fragment that we analysed (TTP(Zn)) did not derive from inappropriate folding or other artifactual effects. Firstly, TTP(Zn) included the entire 77 amino acid TTP fragment that binds and acts on the TNFalpha mRNA (Lai et al., 2000). In addition, the presence of the TTP N- or C-terminal region along with the zinc fingers partially restored apoptotic activity (Figure 2b,d). Finally, the TTP(Zn-C) mutant robustly induced apoptosis, indicating that it is functional, but it did not sensitize cells to TNFalpha-induced apoptosis (Figure 2a,b, 4c,d). Our findings strongly suggest that regions of TTP outside of the zinc fingers are likely to be important for its functions in vivo (Figure 1a).

Both the N- and C-terminal regions of TTP influence its localization within the cell. TTP(Zn) was present predominantly in the nucleus, and the N- and C-terminal regions were each required for normal localization of TTP to the cytoplasm (Figure 3). After submission of this manuscript, it was reported that the entire C-terminal region of TTP is dispensable for its localization to the cytoplasm (Phillips et al., 2002). One possible explanation for this discrepancy is that the latter finding was obtained using GFP-TTP fusion proteins. We have observed that fusion of GFP to its amino terminus abrogates the effects of TTP on TNFalpha responses, and generally enhances the cytoplasmic localization of the various TTP mutants we have analysed (not shown). Apparently, the presence of GFP at its amino terminus interferes with or masks effects of some functional elements within TTP. Because we did not observe a strict correlation between induction of apoptosis and subcellular localization of TTP (Figures 2b and 3a), we conclude that the TTP N- and C-terminal regions are not required simply to localize the TTP zinc fingers, but are involved in interactions or signals that are important for TTP activity. Consistent with this view, expression of TTP zinc finger mutants that lack RNA binding activity stabilizes ARE-containing mRNAs, suggesting that these mutants may act as dominant negatives (Lai et al., 2002). In addition, phosphorylation of TTP within regions outside of the zinc fingers not only influences TTP subcellular localization (Johnson et al., 2002), but also may inhibit its activity in the cell (Carballo et al., 2001; Zhu et al., 2001). The related C. elegans germline protein PIE-1, which also contains two CCCH zinc fingers, provides a precedent for this apparent functional complexity. Distinct but overlapping regions of PIE-1 are required for its subcellular localization, stability, and functions, which include acting in the nucleus to inhibit transcription, and in the cytoplasm to promote translation of particular mRNAs (Batchelder et al., 1999; Reese et al., 2000; Tenenhaus et al., 2001).

The involvement of multiple regions of the TTP protein in its pro-apoptotic effects is reassuring, because it indicates that these effects do not derive simply from forced expression of its RNA binding zinc finger region, and may reflect aspects of its physiological functions. Our findings support the hypothesis that TTP/TIS11 proteins have additional functions besides cytokine mRNA destabilization, even though TTP -/- mice lack obvious defects other than their characteristic inflammatory syndrome (Taylor et al., 1996a). One possible model is that TTP influences cell survival through mechanisms that are largely redundant with other regulatory networks. How might TTP/TIS11 proteins induce apoptosis? Various other CCCH zinc finger proteins also regulate genes post-transcriptionally. In C. elegans, for example, multiple proteins with two characteristic CCCH zinc fingers either promote or inhibit translation (Guedes and Priess, 1997; Schubert et al., 2000; Tabara et al., 1999; Tenenhaus et al., 2001). TTP/TIS11 proteins may similarly influence either stability or translation of mRNAs that are critical for cell growth or survival. In addition, in transfection assays TTP appears to decrease transcription of numerous different genes (Carballo et al., 1998; Zhu et al., 2001). This effect is significantly less prominent in 293 cells however (Carballo et al., 1998), which are resistant to TTP-induced apoptosis (Johnson et al., 2000), raising the question of whether this apparent broad 'squelching' of gene expression might be a cause or a consequence of TTP-induced apoptosis.

TTP induces apoptosis and modulates responses to TNFalpha through distinct mechanisms

TTP is unique among TTP/TIS11 proteins in that it induces apoptosis synergistically with TNFalpha (Johnson et al., 2000). Here we have determined that this synergistic induction of cell death is not inhibited by Bcl-2, which blocks induction of apoptosis by TTP/TIS11 proteins in the absence of TNFalpha (Johnson et al., 2000) (Figure 2a). This finding suggests that TTP and TNFalpha induce apoptosis synergistically independently of the mitochondrial pathway, and implies that TTP sensitizes cells to the direct apoptotic stimulus of TNFalpha (Figure 6). Further supporting the model that TTP induces apoptosis and sensitizes cells to TNFalpha through different mechanisms, although both the TTP(N-Zn) and TTP(Zn-C) mutants can induce apoptosis through the mitochondrial pathway, only TTP(N-Zn) acts synergistically with TNFalpha (Figures 2a, 4 and 6). Since the zinc fingers are relatively conserved among TTP/TIS11 proteins (Figure 1a,c), it appears likely that the more divergent 92 residue N-terminal region of TTP mediates interactions that are essential for its unique influence on TNFalpha responses. We conclude that TTP (but not TIS11b or TIS11d) alters the balance between the survival and death signals induced by TNFalpha (Figure 6), most likely by affecting expression of a component or end-product of these signaling pathways.

Transfection and in vitro studies strongly support the model that TTP limits production of TNFalpha and other cytokines by directly binding and destabilizing their mRNAs (Carballo et al., 1998, 2000; Chen et al., 2001; Lai and Blackshear, 2001; Lai et al., 1999, 2000). TTP may act on the TNFalpha mRNA through more than one mechanism however, because at some expression levels TTP appears to stabilize particular TNFalpha mRNA species (Lai et al., 1999). In addition, although both TIS11b and TIS11d robustly bind and destabilize cytokine mRNAs in transfection assays (Lai et al., 2000), they each fail to sensitize cells to the apoptotic stimulus of TNFalpha (Johnson et al., 2000), indicating that its influence on TNFalpha responses is a highly specific characteristic of TTP.

Individual orthologs of each TTP/TIS11 protein appear to be conserved in vertebrates (Lai et al., 2000), but not in C. elegans and Drosophila, which each encode a single related protein that at some zinc finger positions is more similar to TIS11b or TIS11d (Figure 1c). The two most closely related genes in S. cerevisiae similarly do not each correspond directly to an individual vertebrate TTP/TIS11 gene (Figure 1c). These sequence relationships suggest that the three TTP/TIS11 genes may have arisen from a common ancestor specifically in vertebrates, and that TTP may have diverged in tandem with specialized functions that are reflected by its influence on TNFalpha responses.

The unique effect of TTP on TNFalpha responses could be important in vivo in multiple ways. Maturing activated monocytes face a choice between NF-kappaB-dependent survival and differentiation, and death by TNFalpha-induced apoptosis (Figure 6) (Pennington et al., 2001). If TTP were involved in this decision in certain hematopoietic lineages, its absence could result in some inflammatory responses not being regulated appropriately. In addition, certain effects of TTP are consistent with the speculative but intriguing possibility that TTP might inhibit stress-activated protein kinase pathways. These signaling pathways are induced by TNFalpha to promote cell survival (Figure 6), and they enhance TNFalpha mRNA translation and stability through the ARE (Kontoyiannis et al., 2001; Rutault et al., 2001; Vasudevan and Peltz, 2001; Winzen et al., 1999). By inhibiting these stress-activated signals, TTP could both enhance the apoptotic stimulus of TNFalpha, and decrease TNFalpha mRNA stability or translation indirectly. In addition, it may be significant that the stress-activated p38 pathway is required for induction of TTP expression, but also phosphorylates TTP (Mahtani et al., 2001), and thereby appears to decrease its RNA binding affinity and inhibit its activity in the cell (Carballo et al., 2001; Zhu et al., 2001). These complex relationships suggest that by acting post-transcriptionally on different target mRNAs, TTP may influence multiple interconnected feedback loops. Further elucidation of the mechanisms through which TTP induces apoptosis and modulates TNFalpha responses may yield important insights into how TTP influences inflammatory events in vivo.

Materials and methods


TTP regions shown in Figure 1b were introduced into the CMV-based expression vector CS2+(Turner and Tjian, 1989) by PCR (Pfu, Stratagene), with a Kozak consensus and ATG added where appropriate. An N-terminal GFP fusion was made by restriction cloning full length TTP from CS2TTP into C2eGFP (Clontech). N-terminal fusions of TTP deletion mutants were restriction cloned into C3eGFP.

Transfections and cell death assays

Transfections were carried out as described (Johnson et al., 2000), in 35 mm plates using Lipofectamine or Lipofectamine Plus (Life Technologies) as indicated. Cell death was assayed by co-transfection of a beta-galactosidase reporter plasmid, and examination of cell morphology after X-gal staining at 24 h post-transfection (Johnson et al., 2000). Under a variety of experimental conditions, the numbers of apoptotic cells identified by this method were reproducibly comparable to those detected by scoring Hoechst-stained pyknotic nuclei (not shown) (Johnson et al., 2000).

Western blotting, antibody production and immunofluorescence

Cells were lysed in 1% TritonX-100, 50 mM Tris pH 8, 150 mM NaCl, 1 mM MgCl2, 1 mM DTT, 10% glycerol, CompleteTM protease inhibitors (Roche Pharmaceuticals), 1 mM Na Vanadate and 50 mM NaF (cell lysis buffer). For Western blotting, 100 mug of protein was used per lane (Johnson et al., 2000). Mouse monoclonal anti-GFP (Zymed) was used according to manufacturers directions. The nTTP antibody was previously described as anti-TTP antibody (Johnson et al., 2000). cTTP was raised (Cocalico Biologicals, PA, USA) and affinity purified similarly, using a peptide corresponding to the 21 C-terminal amino acids of TTP, with a cysteine added N-terminally to allow conjugation. These antisera were used at 1/5000 dilution. Immunofluorescence was carried out as described (Johnson et al., 2000), with 200-300 cells per slide counted for determination of TTP localization.


For helpful discussions and critical reading of the manuscript we thank Karl Munger, Grace Gill, Azad Bonni, Valerie Zacny, Michael Yaffe, and Blackwell lab members. Supported by a grant from the NIH to TK Blackwell. (1 R01 CA84418), and an Arthritis Foundation postdoctoral fellowship to BA Johnson.


Bai C, Tolias PP. (1996). Mol. Cell. Biol., 16: 6661-6667. MEDLINE

Baker SJ, Reddy EP. (1996). Oncogene, 12: 1-9. MEDLINE

Barabino SM, Hubner W, Jenny A, Minvielle-Sebastia L, Keller W. (1997). Genes Dev., 11: 1703-1716. MEDLINE

Batchelder C, Dunn MA, Choy B, Suh Y, Cassie C, Shim EY, Shin TH, Mello C, Seydoux G, Blackwell TK. (1999). Genes Dev., 13: 202-212. MEDLINE

Carballo E, Cao H, Lai WS, Kennington EA, Campbell D, Blackshear PJ. (2001). J. Biol. Chem., 276: 42580-42587. MEDLINE

Carballo E, Gilkeson GS, Blackshear PJ. (1997). J. Clin. Invest., 100: 986-995. MEDLINE

Carballo E, Lai WS, Blackshear PJ. (1998). Science, 281: 1001-1005. Article MEDLINE

Carballo E, Lai WS, Blackshear PJ. (2000). Blood, 95: 1891-1899. MEDLINE

Chen CY, Gherzi R, Ong SE, Chan EL, Raijmakers R, Pruijn GJ, Stoecklin G, Moroni C, Mann M, Karin M. (2001). Cell, 107: 451-464. MEDLINE

Chen CY, Shyu AB. (1995). Trends Biochem. Sci., 20: 465-470. MEDLINE

Chen MJ, Holskin B, Strickler J, Gorniak J, Clark MA, Johnson PJ, Mitcho M, Shalloway D. (1987). Nature, 330: 581-583. MEDLINE

DuBois RN, Bishop PR, Graves-Deal R, Coffey RJ. (1995). Cell. Growth Differ., 6: 523-529. MEDLINE

DuBois RN, McLane MW, Ryder K, Lau LF, Nathans D. (1990). J. Biol. Chem., 265: 19185-19191. MEDLINE

Dumitru CD, Ceci JD, Tsatsanis C, Kontoyiannis D, Stamatakis K, Lin JH, Patriotis C, Jenkins NA, Copeland NG, Kollias G, Tsichlis PN. (2000). Cell, 103: 1071-1083. MEDLINE

Ehrenfried JA, Townsend JrCM, Thompson JC, Evers BM. (1995). Ann. Surg., 222: 51-56. MEDLINE

Green DR. (2000). Cell, 102: 1-4. MEDLINE

Guedes S, Priess J. (1997). Development, 124: 731-739. MEDLINE

Haas CA, Donath C, Kreutzberg GW. (1993). Neuroscience, 53: 91-99. MEDLINE

Harkin DP, Bean JM, Miklos D, Song YH, Truong VB, Englert C, Christians FC, Ellisen LW, Maheswaran S, Oliner JD, Haber DA. (1999). Cell, 97: 575-586. MEDLINE

Heximer SP, Forsdyke DR. (1993). DNA Cell. Biol., 12: 73-88. MEDLINE

Janicke RU, Lee FH, Porter AG. (1994). Mol. Cell. Biol., 14: 5661-5670. MEDLINE

Johnson BA, Geha M, Blackwell TK. (2000). Oncogene, 19: 1657-1664. MEDLINE

Johnson BA, Stehn JR, Yaffe MB, Blackwell TK. (2002). J. Biol. Chem. Published on line.

Klefstrom J, Arighi E, Littlewood T, Jaattela M, Saksela E, Evan GI, Alitalo K. (1997). EMBO J., 16: 7382-7392. MEDLINE

Kontoyiannis D, Kotlyarov A, Carballo E, Alexopoulou L, Blackshear PJ, Gaestel M, Davis R, Flavell R, Kollias G. (2001). EMBO J., 20: 3760-3770. MEDLINE

Kontoyiannis D, Pasparakis M, Pizarro TT, Cominelli F, Kollias G. (1999). Immunity, 10: 387-398. MEDLINE

Kotlyarov A, Neininger A, Schubert C, Eckert R, Birchmeier C, Volk HD, Gaestel M. (1999). Nat. Cell. Biol., 1: 94-97. Article MEDLINE

Lai WS, Blackshear PJ. (2001). J. Biol. Chem., 276: 23144-23154. MEDLINE

Lai WS, Carballo E, Strum JR, Kennington EA, Phillips RS, Blackshear PJ. (1999). Mol. Cell. Biol., 19: 4311-4323. MEDLINE

Lai WS, Carballo E, Thorn JM, Kennington EA, Blackshear PJ. (2000). J. Biol. Chem., 275: 17827-17837. MEDLINE

Lai WS, Kennington EA, Blackshear PJ. (2002). J. Biol. Chem., 277: 9606-9613. MEDLINE

Lai WS, Stumpo DJ, Blackshear PJ. (1990). J. Biol. Chem., 265: 16556-16563. MEDLINE

Landesberg LJ, Ramalingam R, Lee K, Rosengart TK, Crystal RG. (2001). Am. J. Physiol. Lung Cell. Mol. Physiol., 281: L1138-L1149. MEDLINE

Mahtani KR, Brook M, Dean JL, Sully G, Saklatvala J, Clark AR. (2001). Mol. Cell. Biol., 21: 6461-6469. MEDLINE

Mesner PW, Epting CL, Hegarty JL, Green SH. (1995). J. Neurosci., 15: 7357-7366. MEDLINE

Ming XF, Stoecklin G, Lu M, Looser R, Moroni C. (2001). Mol. Cell. Biol., 21: 5778-5789. MEDLINE

Mittelstadt PR, DeFranco AL. (1993). J. Immunol., 150: 4822-4832. MEDLINE

Nakajima K, Wall R. (1991). Mol. Cell. Biol., 11: 1409-1418. MEDLINE

Pennington KN, Taylor JA, Bren GD, Paya CV. (2001). Mol. Cell. Biol., 21: 1930-1941. MEDLINE

Phillips AC, Ernst MK, Bates S, Rice NR, Vousden KH. (1999). Mol. Cell., 4: 771-781. MEDLINE

Phillips R, Ramos S, Blackshear PJ. (2002). J. Biol. Chem., 277: 11606-11613. MEDLINE

Raghavan A, Robison RL, McNabb J, Miller CR, Williams DA, Bohjanen PR. (2001). J. Biol. Chem., 276: 47958-47965. MEDLINE

Ray CA, Black RA, Kronheim SR, Greenstreet TA, Sleath PR, Salvesen GS, Pickup DJ. (1992). Cell, 69: 597-604. MEDLINE

Reese KJ, Dunn MA, Waddle JA, Seydoux G. (2000). Mol. Cell., 6: 445-455. MEDLINE

Rutault K, Hazzalin CA, Mahadevan LC. (2001). J. Biol. Chem., 276: 6666-6674. MEDLINE

Schubert CM, Lin R, de Vries CJ, Plasterk RH, Priess JR. (2000). Mol. Cell., 5: 671-682. MEDLINE

Tabara H, Hill RJ, Mello CC, Priess JR, Kohara Y. (1999). Development, 126: 1-11. MEDLINE

Taylor GA, Carballo E, Lee DM, Lai WS, Thompson MJ, Patel DD, Schenkman DI, Gilkeson GS, Broxmeyer HE, Haynes BF, Blackshear PJ. (1996a). Immunity, 4: 445-454. MEDLINE

Taylor GA, Thompson MJ, Lai WS, Blackshear PJ. (1995). J. Biol. Chem., 270: 13341-13347. MEDLINE

Taylor GA, Thompson MJ, Lai WS, Blackshear PJ. (1996b). Mol. Endocrinol., 10: 140-146.

Tenenhaus C, Subramaniam K, Dunn MA, Seydoux G. (2001). Genes Dev., 15: 1031-1040. MEDLINE

Turner R, Tjian R. (1989). Science, 243: 1689-1694. MEDLINE

Varnum BC, Lim RW, Sukhatme VP, Herschman HR. (1989). Oncogene, 4: 119-120. MEDLINE

Varnum BC, Ma QF, Chi TH, Fletcher B, Herschman HR. (1991). Mol. Cell. Biol., 11: 1754-1758. MEDLINE

Vasudevan S, Peltz SW. (2001). Mol. Cell., 7: 1191-1200. MEDLINE

Winzen R, Kracht M, Ritter B, Wilhelm A, Chen CY, Shyu AB, Muller M, Gaestel M, Resch K, Holtmann H. (1999). EMBO J., 18: 4969-4980. Article MEDLINE

Zhu W, Brauchle MA, Di Padova F, Gram H, New L, Ono K, Downey JS, Han J. (2001). Am. J. Physiol. Lung Cell. Mol. Physiol., 281: L499-L508. MEDLINE

Zuo P, Maniatis T. (1996). Genes Dev., 10: 1356-1368. MEDLINE


Figure 1 Mutagenesis analysis of TTP. (a) Diagram showing the percentage identity between different regions of mouse TTP/TIS11 proteins and human TTP (hTTP, or GOS24 (Heximer and Forsdyke, 1993)). The CCCH zinc finger regions are represented in red. Blue and orange denote the N and C terminal regions respectively. (b) TTP mutants. Red boxes indicate the CCCH zinc fingers, and each cross in mutant M1,2 indicates the substitution of a critical zinc finger cysteine with serine (Johnson et al., 2000). Blue and orange boxes represent the 23 and 21 amino acids at the TTP N- and C-termini, respectively, against which anti-peptide antibodies were raised (nTTP and cTTP). Amino acids present in each construct are indicated in parentheses, except that TTP(N-Zn) and TTP(Zn-C) also include the 21 C-terminal amino acids to allow detection with cTTP. (c) Comparison of the zinc finger regions of mouse TTP/TIS11 proteins (TTP, TIS11b and TIS11d; Genbank accession numbers P22893, P23950 and P23949, respectively) with those of their closest relatives in D. melanogaster (DTIS11; P47980), C. elegans (F38b7.1; T21955) and S. cerevisiae (CTH1; S57977 and CTH2; P47977). Sequence similarities to mouse TIS11b (Clustal method) are boxed. Signature residues that define vertebrate TTP, TIS11b, and TIS11d proteins (Lai et al., 2000) are shown in red, purple and green, respectively. Positions are numbered as in the full-length proteins

Figure 2 Multiple regions of TTP are necessary to induce apoptosis. (a) Inhibition of TTP-induced cell death by Bcl-2 and CrmA. Cells were transfected with 200 ng of constructs that expressed the indicated TTP mutants (Figure 1b), along with 100 ng of beta-gal reporter and either 200 ng of Bcl-2 or 1.7 mug of CrmA vector, using Lipofectamine and 2 mug total DNA. After 24 h, cells were X-gal stained and the percentage of apoptotic blue cells was determined. Bars represent the mean of four wells and error bars represent the standard deviation. (b) Apoptosis induced by TTP mutants. 3T3 cells were transfected and assayed for apoptosis as in a. (c) Western blot showing TTP expression levels in a duplicate of the experiment in b. Molecular weights and a background band are indicated to the left of the gel. cTTP and nTTP indicate the antibodies used for detection. Lanes 7 and 8 are derived from the same lysates as lanes 1 and 2, respectively. (d) Induction of apoptosis by GFP-tagged TTP proteins (green), compared to their untagged counterparts (red), assayed as in b. Each bar represents the mean of 2 wells. (e) TTP expression in a duplicate of the experiment in d, assayed using cTTP and GFP antibodies as in c. The lower GFP-TTP(Zn) band in lane 2 (bottom) presumably corresponds to a breakdown product

Figure 3 Subcellular localization of TTP mutants. (a) 3T3 cells were transfected as in Figure 2a with 200 ng of the indicated TTP mutant expression vector (Figure 1b), along with 200 ng of Bcl-2 vector to prevent apoptosis, then after 24 h were fixed and stained with the indicated TTP antibody. A representative experiment is shown. The subcellular localization of these various TTP mutants was generally comparable over a wide range of expression levels (not shown). (b) Representative cells in which TTP mutants were localized as classified in a, at 100´magnification. Antibodies used for staining are indicated to the right, and a box surrounds views of two fields from the same TTP sample. Endogenous TTP is not detectable in these continuously dividing cells (not shown)

Figure 4 The TTP N-terminal and zinc finger regions are required to induce apoptosis synergistically with TNFalpha. (a) Both TTP and TTP(N-Zn) act synergistically with TNFalpha. A representative experiment in which 3T3 cells were transfected with 25 ng of the indicated TTP expression or CS2 control vector along with 100 ng of beta-gal reporter plasmid, using Lipofectamine Plus and 1 mug total DNA. TNFalpha was added to the indicated concentrations after 3 h, then after an additional 21 h cell death was assayed as in Figure 2a. Each point is the mean of four wells in a typical experiment. The increase in mean percentage apoptosis associated with exposure to 1 ng ml-1 TNFalpha is defined as 'synergy,' and is indicated on the right for each transfected protein. For TTP, this apoptosis synergy varied between 15 and 33% over several experiments (not shown). (b) Rapid synergistic induction of apoptosis by TTP and TNFalpha when protein synthesis is inhibited. Twenty-four hours after cells were transfected as described in a, they were incubated in TNFalpha and cycloheximide (10 mug ml-1) for 4 h, then assayed for apoptosis. (c) TTP(Zn) and TTP(Zn-C) do not promote apoptosis synergistically with TNFalpha, even when expressed at higher levels. Transfections were performed and analysed as in a, but with the indicated amounts of expression plasmid. (d) Increase in apoptosis associated with addition of TNFalpha, determined as in a, for each transfection shown in c

Figure 5 TTP sensitizes cells to the apoptotic stimulus of TNFalpha independently of the mitochondrial pathway. (a) TTP promotes Bcl-2-resistant apoptosis in response to TNFalpha. Transfections were performed and assayed for cell death as in Figure 4a, except that either 100 ng of Bcl-2 or 875 ng of CrmA expression vector was added as indicated. (b) Apoptotic synergy between TNFalpha and TTP is inhibited by CrmA, but not Bcl-2. The percentage increase in apoptosis associated with TNFalpha addition is graphed for each data set described in a (CS2, TTP, TTP(N-Zn))

Figure 6 Diagram illustrating how TTP influences apoptosis signaling pathways. TTP induces apoptosis through the mitochondrial pathway, which is inhibited by Bcl-2 (Johnson et al., 2000). This effect requires that either the TTP N- or C-terminal regions be present in addition to the zinc fingers, and is reduced by deletion of either of these regions (Figure 2). When TTP is expressed at lower levels, treatment with modest concentrations of TNFalpha induces apoptosis that is resistant to Bcl-2 but sensitive to CrmA (Figure 5), implying that this cell death derives from direct apoptotic signaling by TNFalpha (Green, 2000). This effect of TTP on the response to TNFalpha specifically requires the TTP N-terminal region along with the zinc fingers (N-Zn; Figure 4). Grey lines indicate that TTP may act downstream of TNFalpha either to (a) inhibit survival signals, or (b) enhance direct apoptotic signals

Received 14 December 2001; revised 18 March 2002; accepted 26 March 2002
20 June 2002, Volume 21, Number 27, Pages 4237-4246
Table of contents    Previous  Article  Next    [PDF]