Tissue Transglutaminase Protects against Apoptosis by Modifying the Tumor Suppressor Protein p110 Rb*

Tissue transglutaminase (TGase) is involved in the regulation of several biological events including cellular differentiation and apoptosis. The expression and activation of TGase are up-regulated in response to retinoic acid (RA), leading to the protection of several cell lines againstN-(4-hydroxyphenyl)retinamide (HPR)-induced apoptosis. The anti-apoptotic mechanisms of TGase are poorly understood at this time. We examined the interaction of TGase with the retinoblastoma (Rb) protein, a substrate of TGase that is also implicated in cell survival functions. In cells undergoing HPR-induced apoptosis, Rb is degraded. This degradation is blocked when cells are pretreated with RA, an important regulator of TGase. In vitro studies revealed that TGase protects Rb from caspase-induced degradation in a transamidation-dependent manner. Experiments performed with fibroblasts from Rb−/− mice further demonstrated that the presence of Rb was required for TGase to exhibit anti-apoptotic activity in response to RA treatment. Microinjection of Rb−/− cells with a transamidation-defective TGase mutant and Rb afforded no protection from HPR-induced apoptosis. Taken together, these findings suggest that the ability of TGase to modify Rb via transamidation underlies the ability of TGase to provide protection against apoptotic insults and to ensure that cells remain viable during differentiation.

Tissue transglutaminase or type II transglutaminase (TGase) 1 is an 87-kDa protein that contains two key catalytic activities, the ability to catalyze protein-amine cross-links and a GTP binding and hydrolysis activity. The transamidation reaction of TGase has been well studied and consists of the Ca 2ϩ -dependent formation of covalent bonds between the ␥-carboxamide groups of peptide-bound glutamine residues and the primary amino groups of a wide variety of proteins (1,2). Transamidation has been implicated in a number of biological processes such as axonal regeneration, cellular differentiation, and apoptosis (3)(4)(5)(6)(7)(8)(9)(10)(11)(12). Many of the early studies of TGase have identified it as being present in cells and tissues undergoing apoptosis (7-10, 13, 14) and implicated the transamidation activity of TGase as a potentiator of programmed cell death (15,16). However, there is growing evidence that TGase may not directly mediate apoptosis. TGase Ϫ/Ϫ mice showed no major developmental abnormalities, and the thymocytes from these mice were no less susceptible to apoptosis than TGase ϩ/ϩ cells (17,18). Studies examining the role of TGase in retinoic acid (RA)-mediated signaling (12), as well as studies demonstrating that TGase was required for neurite outgrowth (19), suggest that TGase may exhibit protective effects against apoptotic signals. It has been well documented that RA up-regulates both the expression and transamidation activity of TGase (12, 20 -22). We have recently shown that the ability of RA to up-regulate TGase expression and activity is required for the ability of RA to protect against apoptosis induced by a synthetic retinoid analog, all-trans-N-(4-hydroxyphenyl)retinamide (HPR) (12). The TGase-mediated inhibition of apoptosis appeared to involve its transamidation activity and its ability to bind GTP (12), a result that may suggest a role for TGase in more than one anti-apoptotic signaling pathway.
To understand how TGase is acting as an anti-apoptotic factor, we set out to investigate substrates of TGase that are potential regulators of cell death. One such protein was the retinoblastoma gene product (Rb), a well established regulator of the G 1 /S checkpoint in the cell cycle (23)(24)(25). In addition to its role in cell cycle regulation and cellular differentiation, Rb has also been implicated in apoptosis (26). Rb Ϫ/Ϫ mice are embryonic lethal with significant apoptosis occurring in the developing nervous system and lens of the eye (27)(28)(29)(30). The tumor suppressor gene p53 activates a pathway in response to various cellular insults that results in the degradation of Rb (31), a critical event that shifts cells toward apoptosis. The Rb protein was initially identified as a substrate for TGase-mediated transamidation in U937 cells in early stages of apoptosis (11). This finding, when combined with other results showing increased expression of TGase in apoptotic cells (7-10, 13, 14), led to the suggestion that TGase was a pro-apoptotic protein. However, we have found that TGase acts in a protective fashion (12) by modifying proteins during times of cellular stress, thereby creating a window for survival against apoptotic stimuli and enabling cells to continue to grow or differentiate. We show here that the transamidation activity of TGase protects Rb against caspase-catalyzed degradation and that this protective effect is crucial for the anti-apoptotic actions of TGase.

EXPERIMENTAL PROCEDURES
Cell Culture-HL60 cells were maintained in RPMI 1640 medium (Cellgro) with 10% heat-inactivated fetal calf serum (Atlanta Biologicals), 2 mM glutamine, 100 IU/ml penicillin, 100 g/ml streptomycin, 2 g/liter sodium bicarbonate, and 10 mM HEPES in a 37°C incubator with 5% CO 2 . Mouse embryonic lung fibroblasts were grown in low glucose Dulbecco's modified Eagle's medium (Invitrogen) with 10% fetal bovine * This work was supported by National Institutes of Health Grant GM61762. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. serum (Atlanta Biologicals) and 100 IU/ml penicillin, 100 g/ml streptomycin, 2 g/liter sodium bicarbonate, and 10 mM HEPES in a 37°C incubator with 5% CO 2 . Upon stimulation with 5 M RA (Sigma) or 5 M HPR (Sigma), the cells were switched to serum-free medium.
Purification of Recombinant Proteins-Baculovirus-expressed Rb was purified using E7-peptide (TDLYCYEQLN)-coated Sepharose. Briefly, Sf21 cells infected with Rb were pelleted and resuspended in 1.5 ml of HMGN buffer (250 mM NaCl, 5 mM ␤-mercaptoethanol, 0.2 mM phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, and 10 g/ml aprotinin) and sonicated four times for 15 s at half-maximal power on ice. The lysate was centrifuged at 4°C, and the supernatant was transferred to a 1.5-ml Eppendorf tube containing 65 l of E7-peptide-coated Sepharose in HMGN buffer. The mixture was rotated for 30 min at 4°C and pelleted in a microcentrifuge at 4°C, and the beads were washed (three times) with 500 l of HMGN buffer. Elution buffer (100 l) (50 mM Na 2 CO 3 , 150 mM NaCl, 1 mM dithiothreitol, and 0.2 mM phenylmethylsulfonyl fluoride) was added to the beads and incubated on ice for 30 min. Samples were centrifuged for 30 s at 4°C, and the supernatant was placed in a new Eppendorf tube and supplemented with 4 l of HEPES, pH 7.4, and 24 l of 50% glycerol. The elution process was repeated, and the eluate was pooled and stored at Ϫ80°C. GST-TGase was purified as described previously (32).
In Vitro Caspase Cleavage Assay-TGase-mediated protection of Rb was assayed by incubating insect cell-expressed Rb (10.5 M) with GST-TGase (10.5 M) in 25 l of transamidation reaction buffer (25 mM HEPES, pH 7.4, 100 mM NaCl, 5 mM KCl, 0.3 mM Na 2 HPO 4 , 1 mM NaHCO 3 , 5 mM glucose, 20 mM dithiothreitol, 0.8% glycerol, 1 mM CaCl 2 , and 250 M putrescine). For samples serving as negative controls, CaCl 2 was omitted. Reactions were incubated at 25°C for 30 min. Upon completion of the transamidation reaction, the samples were supplemented with 25 l of caspase cleavage buffer (100 ng of caspase-7, 4 mM EDTA, 0.2% CHAPS, and 20% sucrose) for a total reaction volume of 50 l. The samples were incubated at 37°C for 1, 5, 15, and 60 min. Reactions were stopped by the addition of 2ϫ loading buffer, followed by boiling for 5 min. Reaction mixtures were resolved on 8% polyacrylamide gels and analyzed by Western blot as described above.
Apoptotic Assays-5 ϫ 10 4 ELF Rb ϩ/ϩ and Rb Ϫ/Ϫ cells were plated onto flame-sterilized coverslips. The next day, cells were incubated in serum-free medium, and apoptosis was induced by the addition of 5 M HPR for 18 h. Post-treatment, the cells were washed twice with phosphate-buffered saline (2.7 mM KCl, 1.5 mM KH 2 PO 4 , 137 mM NaCl, and 8.1 mM Na 2 HPO 4 ) and stained with 2.5 g/ml Hoechst 33258. At least 500 nuclei were counted for each treatment, and the percentage of apoptotic nuclei (condensed and blebbing) was calculated. Apoptosis in cells microinjected with Rb and TGase constructs was determined by counting the number of GFP-expressing cells prior to the addition of HPR (70 -200 cells of 300 injected successfully). Ten hours after treatment with HPR, the surviving cells expressing GFP were counted, and the percentage of cell death was determined. Cell death caused by microinjection alone was controlled by calculating the cell death incurred by cells injected with GFP alone and left untreated for 10 h. This percentage was subtracted from the percentage of cell death induced by a particular condition. Each experimental condition was repeated independently at least three times.

RESULTS AND DISCUSSION
Tissue transglutaminase (TGase) has been shown to be upregulated in cells undergoing apoptosis, raising the possibility that it is a pro-apoptotic protein (7)(8)(9)(10)(13)(14)(15)(16). However, recent studies examining the role of TGase in RA-mediated differentiation have shown that it provides protection against apoptosis (12) and may be important for cellular differentiation (19,33). To better understand the anti-apoptotic activity of TGase, we wanted to examine its interactions with substrates that may be involved in the regulation of apoptosis. One substrate that was particularly interesting was the 110-kDa Rb. This well known regulator of the cell cycle (23,24) and cellular differentiation (25,26) was identified as a target for TGase in U937 cells undergoing the early stages of apoptosis (11). Rb is a target of caspase degradation in p53-induced apoptosis (31), and Rb-knockout mice are embryonic lethal and exhibit large amounts of apoptosis in the developing retina and nervous system (27)(28)(29)(30).
To determine whether TGase interacted with Rb in a manner that would be protective to cells, we first examined the effects of RA and HPR on Rb in NIH 3T3 cells. It is well documented that RA is a regulator of TGase expression (12, 20 -22), and we have shown that the ability of RA to up-regulate TGase is important for cell survival against an apoptotic challenge as induced by the chemotherapeutic retinoid analog HPR (12). Whole cell lysates from HL60 cells treated with either 5 M RA (0 -72 h) or 5 M HPR (0 -18 h) were subjected to SDS-PAGE and analyzed by Western blot for the presence of Rb. In cells treated with RA, the Rb protein is clearly present and appears unchanged (Fig. 1A), whereas in cells treated with HPR, the cleaved form of Rb begins to appear just 6 h post-treatment and by 18 h it appears that nearly all of the cellular Rb has been degraded (Fig. 1A). Although these findings did not necessarily indicate a direct role for TGase in protecting Rb from degradation, they did demonstrate that the degradation of Rb is a specific event in the process of HPR-induced apoptosis. In cells pretreated with RA for 2 days, we were able to inhibit apoptosis induced by HPR, a result dependent on up-regulation of TGase expression and activity (12). In cells pretreated with RA for 2 days, prior to treatment with HPR, we observed that Rb was no longer degraded (Fig. 1B). The inhibition of HPR-induced Rb degradation by RA may be due to the up-regulation and increase in TGase activity that is a direct outcome of RA stimulation. Indeed, a model for TGase-mediated protection of Rb from caspase degradation is quite attractive, considering that 5 glutamine residues that may act as amine donors for TGasecatalyzed transamidation flank the caspase cleavage site on Rb. It is possible that modification of one or more of these To test whether TGase is directly involved in preventing the caspase-induced degradation of Rb, as well as determine if the transamidation reaction of TGase is required, we looked at the ability of recombinant TGase to protect recombinant Rb from caspase-7 cleavage using a well defined assay system. Rb was incubated with or without TGase in the presence of putrescine, a polyamine that can serve as an amine acceptor for the transamidation reaction. The reactions were carried out in the presence or absence of CaCl 2, an essential cofactor for the transamidation reaction catalyzed by TGase (1,2). After incubation of TGase with Rb, activated caspase-7 was added for 1, 5, 15, and 60 min, and the reactions were resolved by SDS-PAGE and blotted with an anti-Rb antibody specific for the amino-terminal portion of Rb (allowing the detection of both the full-length and caspase-cleaved forms of Rb). In reactions containing only Rb and TGase, we did not see the appearance of any cleavage products (Fig. 2, lanes 12  and 13). In mixtures containing Rb, TGase, and caspase-7, but lacking calcium, the cleaved form of Rb appeared within 1 min of exposure to caspase-7 (Fig. 2, lane 3). After 60 min of incubation with caspase-7 in the absence of calcium, Rb was completely degraded (Fig. 2, lane 10). However, when the reactions were performed in the presence of calcium, the ability of caspase-7 to degrade Rb was greatly reduced. Even after 1 h of exposure to the caspase, the majority of the Rb that was incubated in the presence of TGase and calcium was still in its native form (Fig. 2, lane 11). When caspase-7 was incubated with Rb in the absence of TGase for 5 min, significant degradation of Rb was observed (Fig. 2, lane 7). Because the TGase-mediated protection of Rb was CaCl 2 -dependent, we can infer that the TGase-catalyzed transamidation activity underlies this protection rather than some type of complex formation between TGase and Rb. These results strongly support the idea that the ability of TGase to modify Rb via transamidation is crucial for the protection of Rb from caspase cleavage and may play a major role in the ability of TGase to protect cells against apoptosis.
To determine if the interaction between TGase and Rb was indeed important for TGase-mediated survival, we tested the ability of TGase to protect against HPR-induced apoptosis in cells lacking Rb. ELF cells from either Rb Ϫ/Ϫ or wild-type (Rb ϩ/ϩ ) mice were treated with or without RA from 0 to 3 days. As in other cell types, RA stimulated an increase in TGase expression in both the wild-type and the Rb Ϫ/Ϫ cells (Fig. 3A). If the ability of TGase to protect cells from apoptosis was linked to its interaction with Rb, one might expect that only the wild-type Rb ϩ/ϩ cells would be resistant to HPR-induced apoptosis in the presence of RA. To investigate the ability of TGase to protect these cells from apoptosis, both wild-type and Rb Ϫ/Ϫ cells were pretreated for 48 h with 5 M RA and then treated with or without 5 M HPR in the presence or absence of the competitive TGase inhibitor MDC. When the wild-type fibroblasts were pretreated with RA and then exposed to HPR, only 21.4% of the cells underwent apoptosis, which is a significantly lower (p ϭ 0.0061) amount of cell death compared with the percentage of cells (41.6%) undergoing apoptosis when wild-type cells were treated with HPR alone (Fig. 3B). Upon adding the competitive inhibitor MDC, 40.9% of the wild-type cells underwent apoptosis, an extent similar to that for cells treated with HPR alone (Fig. 3B), thereby demonstrating that by inhibiting the transamidation activity of TGase there is a loss of protection by RA. In the Rb Ϫ/Ϫ cells, pretreatment with RA did not protect the cells from apoptosis as 38.3% of the cells were dead after 12 h of exposure to HPR (Fig. 3B), a value not statistically different (p ϭ 0.199) from the 48.7% of Rb Ϫ/Ϫ cells that underwent apoptosis after a 12-h exposure to HPR alone. These results suggest that the anti-apoptotic actions of TGase may rely on its ability to transamidate Rb.
To demonstrate that the combination of TGase activity and Rb is required for the protective effects by TGase, Rb Ϫ/Ϫ fibroblasts were microinjected with DNA encoding either the wildtype TGase or a catalytically inactive form of TGase (TG(C277V)) or a truncated version of TGase that lacked the most carboxyl-terminal ␤-barrel (TG⌬C1), in combination with DNA encoding Rb and GFP (as a marker for injection). Eighteen hours after injection, the cells were treated with or without 5 M HPR for 8 h, and the percentage of injected cells undergoing apoptosis was determined. Cells that either were not injected or were injected with GFP alone, showed high levels of apoptosis when treated with HPR (63.6% of the uninjected cells were apoptotic, and 55.5% of the cells injected with GFP were apoptotic (Fig. 4)). Cells injected with Rb alone also showed relatively high levels of apoptosis when treated with HPR (Fig.  4). Interestingly, cells injected with Rb and wild-type TGase, when treated with HPR, showed a significantly lower level of apoptosis with only 28.8% of the cells undergoing programmed cell death. Cells injected with the transamidation-defective TGase mutant, TG(C277V), and Rb were afforded no protection and exhibited high levels of apoptosis (47.8%), comparable with  4, 6, 9, and 11) or absence (lanes 1, 3, 5, 8, and 10) of 1 mM Ca 2ϩ from 0 to 60 min. The reactions were subjected to SDS-PAGE, transferred to polyvinylidene difluoride membranes, and probed with antibodies to the amino terminus of Rb to detect the presence of the intact and caspase-cleaved forms of Rb. the levels measured in uninjected cells treated with HPR (Fig.  4). Cells injected with the TG⌬C1 mutant also showed significantly elevated levels of apoptosis (41.2%) when treated with HPR (Fig. 4). This is interesting as the carboxyl-terminal ␤-barrel that is missing from this TGase-deletion mutant may be necessary for binding Rb. 2 Overall, these results provide strong evidence that the ability of TGase to interact with and modify Rb via transamidation plays a critical role in the antiapoptotic mechanism of TGase.
The fact that TGase-mediated protection in these cells relies heavily on its transamidation activity is interesting, given that our previous work showed that the catalytically defective TG(C277V) mutant was still able to protect NIH3T3 cells from HPR-induced apoptosis. However, the results presented here, as well as our earlier work showing that MDC blocked protection by TGase, argues that transamidation activity is necessary for protection in certain cell types, especially in mouse embryonic fibroblasts. It is possible that the requirement for transamidation activity in some instances of TGase-mediated protection and the requirement for GTP binding activity in others is specific to cell type. The embryonic fibroblasts have not been exposed to the complete profile of developmental signals that stimulate the growth and differentiation of some cells and the death of other cells. Thus, on a developmental level, the Rb Ϫ/Ϫ cells are at a stage where the presence of Rb is critical for these life or death "decisions." This would account for the need of Rb to be protected by TGase in order for these cells to survive an apoptotic stimuli, whereas other cell types, such as NIH3T3 cells, are no longer completely dependent upon a survival pathway regulated by Rb and instead are more capable of activating alternative survival pathways, for example, one linked to phosphatidylinositol 3-kinase (34). The fact that transamidation activity is important for the protection afforded by TGase in some cell types does not eliminate the possibility that the GTP binding activity of TGase also contributes to its protective effects. Rather, it may reflect a matter of timing where in some instances it is more important for the cells' ultimate survival to modify specific proteins like Rb before activating other antiapoptotic signaling pathways.
In summary, our results demonstrate that an interaction between TGase and Rb, which requires a transamidation-competent TGase, is necessary for the ability of TGase to inhibit apoptosis. These findings would argue for the actions of TGase as an anti-apoptotic factor. It is quite possible that the upregulation of TGase often observed in cells undergoing apoptosis (7-10, 13, 14) represents a cellular regulatory mechanism to block or delay the onset of cell death, rather than reflecting a direct participation by TGase in programmed cell death. Apoptotic signaling pathways are tightly regulated, and there are often several levels of regulation necessary for protection against cell death. By modifying proteins like Rb, TGase may extend the lifetime of the cell, slowing the process of programmed cell death and allowing the appropriate anti-apoptotic pathways to become activated. We are currently working on identifying the sites on Rb that are modified by TGase with the expectation that we will find a pattern or motif common to other targets of TGase involved in anti-apoptotic signaling.