The serine/threonine phosphatase inhibitor calyculin A induces rapid degradation of IκBβ. Requirement of both the N- and C-terminal sequences

Signal-initiated activation of the transcription factor NF-κB is mediated through proteolysis of its cytoplasmic inhibitory proteins IκBα and IκBβ. While most NF-κB inducers trigger the degradation of IκBα, only certain stimuli are able to induce the degradation of IκBβ. The degradation of IκBα is targeted by its site-specific phosphorylations, although the mechanism underlying the degradation of IκBβ remains elusive. In the present study, we have analyzed the effect of phosphatase inhibitors on the proteolysis of IκBβ. We show that the serine/threonine phosphatase inhibitor calyculin A induces the hyperphosphorylation and subsequent degradation of IκBβ in both human Jurkat T cells and the murine 70Z-3 preB cells, which is associated with the nuclear expression of active NF-κB. The calyculin A-mediated degradation of IκBβ is further enhanced by the cytokine tumor necrosis factor-α (TNF-α), although TNF-α alone is unable to induce the degradation of IκBβ. Mutational analyses have revealed that the inducible degradation of IκBβ induced by calyculin A, and TNF-α requires two N-terminal serines (serines 19 and 23) that are homologous to the inducible phosphorylation sites present in IκBα. Furthermore, the C-terminal 51 amino acid residues, which are rich in serines and aspartic acids, are also required for the inducible degradation of IκBβ. These results suggest that the degradation signal of IκBβ may be controlled by the opposing actions of protein kinases and phosphatases and that both the N- and C-terminal sequences of IκBβ are required for the inducible degradation of this NF-κB inhibitor.

The NF-B/Rel family of transcription factors play a pivotal role in the regulation of various cellular genes involved in the immediate early processes of immune, acute phase, and inflammatory responses (1,2). In addition, these cellular factors have also been implicated in the transcriptional activation of certain human viruses, most notably the type 1 human immune deficiency virus (3)(4)(5)(6)(7). The mammalian NF-B/Rel family is composed of at least five structurally related DNA-binding proteins, including p50, p52, RelA, RelB, and c-Rel, which bind to a target DNA sequence (B) as various heterodimers or ho-modimers (reviewed in Siebenlist et al. (8)). In most cell types, including resting T cells, the NF-B/Rel proteins are sequestered in the cytoplasmic compartment by physical association with inhibitory proteins that are characteristic of the presence of various numbers of ankyrin-like repeats (reviewed in Verma et al. (9)). The major cytoplasmic inhibitors include IB␣ (10,11), IB␤ (12), and the precursor proteins of p50 and p52 (9). The IB molecules appear to bind to and mask the nuclear localization signal of NF-B/Rel, thereby preventing the nuclear translocation of these transcription factors (13)(14)(15)(16).
The latent cytoplasmic NF-B/Rel complexes can be activated by a variety of cellular stimuli, including the mitogen phorbol esters, cytokines such as tumor necrosis factor-␣ (TNF-␣), 1 and interleukin-1, the bacterial component lipopolysaccharide, serine/threonine phosphatase inhibitors such as okadaic acid and calyculin A, and the Tax protein from the type I human T cell leukemia virus (HTLV-I) (8,17). Activation of NF-B by these various inducers involves phosphorylation of IB␣ at serines 32 and 36 (18 -23), which in turn targets this inhibitory protein for ubiquitination and proteasome-mediated proteolysis (24,25). Since the IB␣ gene is positively regulated by the NF-B/Rel factors, the depleted IB␣ protein pool can be rapidly replenished through de novo protein synthesis following the activation of NF-B/Rel (26 -31). Thus, IB␣ regulates the transient nuclear expression of NF-B/Rel. Unlike IB␣, IB␤ appears to respond to only certain cellular stimuli, such as lipopolysaccharide, interleukin-1, and Tax, that are known to induce sustained nuclear expression of NF-B/Rel (12,32,33). The depleted IB␤ protein is not immediately resynthesized, which is likely the molecular basis of persistent activation of NF-B/Rel.
The molecular mechanism underlying the differential signal responses between IB␣ and IB␤ remains elusive. Although IB␤ contains two N-terminal serines (serines 19 and 23) that are homologous to the inducible phosphorylation sites of IB␣ (11,12), it is unclear whether phosphorylation can target IB␤ for degradation. Evidence supporting a role of phosphorylation in IB␤ degradation is provided by site-mutagenesis studies which demonstrate that mutation of serines 19 and 23 to alanines abolishes the inducible degration of IB␤ (21,33). In the present study, we have further investigated the role of phosphorylation in the inducible degradation of IB␤ by examining the effect of a serine/threonine phosphatase inhibitor, calyculin A, on the fate of IB␤. We demonstrate that incubation of Jurkat T cells or 70Z/3 pre-B cells with calyculin A is sufficient to induce the hyperphosphorylation and subsequent degradation of IB␤.

Cell Culture and Reagents-Jurkat T cells (ATCC) and
Jurkat cells expressing the SV40 large T antigen (Jurkat Tag) (34) were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, and antibiotics. Murine 70Z/3 pre-B cells (ATCC) were maintained in the same medium supplemented with 50 M ␤-mercaptoethanol. The serine/threonine phosphatase inhibitor calyculin A was purchased from LC Laboratories (Woburn, MA). The proteasome inhibitor MG132 was purchased from ProScript, Inc. (Cambridge, MA). The antibody against the influenza hemagglutinin (HA) epitope tag (anti-HA) was obtained from Boehringer Mannheim. Anti-IB␤ (C-20) was purchased from Santa Cruz Biotechnology, Inc.
Plasmid Constructs and Transient Transfection-The wild type of pCMV4HA-IB␤ was constructed by cloning the IB␤ cDNA (kindly provided by Dr. Sankar Ghosh, Yale University) (12) into a modified pCMV4 expression vector, pCMV4HA (22), downstream of three copies of the HA epitope tag (YPYDVPDYA). IB␤ 19A/23A was generated by substituting serines 19 and 23 with alanines using site-directed mutagenesis (ClonTech, Inc.). IB␤⌬5-27, which lacks amino acids 5-27, was also generated by site-directed mutagenesis. The C-terminal truncation mutant {IB␤(1-308)} was constructed by introducing a stop codon after codon 308 of the wild type IB␤ by restriction digestion (using HindIII), DNA polymerase (Klenow fragment) fill in, and religation. Jurkat Tag cells (5 ϫ 10 6 ) were transfected using DEAE-dextran (35) with the indicated amounts of IB␤ expression vectors. Between 40 and 48 h post-transfection, the cells were incubated with calyculin A (25 nM) and TNF-␣ (10 ng/ml) for the indicated time periods and then subjected to whole extract preparation and immunoblotting analyses as described below.
Immunoblotting and Electrophoresis Mobility Shift Assay (EMSA)-Jurkat cells, 70Z/3 pre-B cells, or transiently transfected Jurkat-Tag cells were stimulated with the indicated inducers and then collected by centrifugation at 800 ϫ g for 5 min. Whole cell and subcellular extracts were prepared as described previously (36,37). For immunoblotting analyses, whole cell extracts (ϳ15 g) were fractionated by reducing 8.75% SDS-PAGE, electrophoretically transferred to nitrocellulose membranes, and then analyzed for immunoreactivity with the indicated primary antibodies using an enhanced chemiluminescence detection system (ECL; Amersham Corp.). For in vitro phosphatase treatment, the extracts were incubated with 20 units of calf intestinal alkaline phosphatase at 35°C for 30 min prior to immunoblotting analysis. EMSA were performed by incubating the nuclear extracts (ϳ 3 g) with a 32 P-radiolabeled high-affinity palindromic B probe, B-pd (coding strand sequence was 5Ј-CAACGGCAGGGGAATTCCCCTCTC-CTT-3Ј) followed by resolving the DNA-protein complexes on native 5% polyacrylamide gels (38).

Calyculin A Induces the Rapid Phosphorylation and Degradation of IB␤ and the Concurrent Nuclear DNA Binding Activity of NF-B in Both Human Jurkat T Cells and Murine 70Z/3 Pre-B Cells-
To investigate the effect of the phosphatase inhibitors on the fate of IB␤, Jurkat T cells were incubated with calyculin A for different time periods followed by analysis of the IB␤ protein by immunoblotting (Fig. 1A, upper panel). In untreated cells, a single 45-kDa form of IB␤ was detected with an IB␤-specific antiserum (Fig. 1A, upper panel,  lane 1). Incubation of the cells with calyculin A (25 nM) for 15 min led to a marked loss of the preexisting IB␤ protein (lane 3), which persisted until at least 1 h after calyculin A stimulation (lanes [3][4][5]. The loss of IB␤ was apparently due to its proteolysis as this effect of calyculin A was blocked by a potent proteasome inhibitor, MG132 (lane 7), known to inhibit the degradation of IB␣ (24,39). Parallel EMSA revealed that degradation of IB␤ was associated with the appearance of the NF-kB DNA binding activity in the nucleus (Fig. 1A, lower  panel). We noticed that the IB␤ isolated from cells treated with calyculin A migrated more slowly on the SDS-polyacrylamide gel compared to the basal form of IB␤ (Fig. 1A, compare  lanes 1 and 6 with lanes 2-5 and 7). To examine whether the slower migration of IB␤ might be due to its phosphorylation, the cell extract was incubated with calf intestine alkaline phosphatase before being subjected to immunoblotting (Fig. 1B).
After calf intestine alkaline phosphatase treatment, the more slowly migrating IB␤ species from calyculin A-treated cells (lane 2) was completely converted to a faster-migrating IB␤ (Fig. 1B, lane 3), thus suggesting that the slower migration of IB␤ in calyculin A-treated cells was due to its phosphorylation. Interestingly, the in vitro dephosphorylated form of IB␤ (lane 3) migrated slightly faster than the basal form present in untreated cells (lane 1). This result suggests that as seen with murine 70Z/3 pre-B cells (40), IB␤ is preexisting in a basally phosphorylated form in untreated Jurkat T cells. This basal form of IB␤ seems to become hyperphosphorylated when the cells are treated with calyculin A. Of note, the hyperphosphorylation of IB␤ appeared to precede its degradation since inhibition of IB␤ degradation by MG132 led to the accumulation of the more slowly migrating hyperphosphorylated IB␤ (Fig. 1A, upper panel, lane 7).
To examine whether the effect of calyculin A on IB␤ could be recapitulated in other cell types, murine 70Z/3 pre-B cells were subjected to the calyculin A treatment. In untreated 70Z/3 cells, IB␤ is preexisting in two forms that migrate with slightly different rates on SDS-PAGE (Fig. 1C, upper panel,  lane 1). The more slowly migrating band was apparently the phosphorylated form of IB␤ since it was converted to the more rapidly migrating form after in vitro incubation with calf intestinal alkaline phosphatase (data not shown) (40). More importantly, incubation of the 70Z/3 cells with calyculin A led to the rapid degradation of the preexisting IB␤ proteins (Fig. 1C,  lanes 2-5, upper panel). Furthermore, as observed in Jurkat T cells, the degradation of IB␤ was preceded by the appearance of the more slowly migrating hyperphosphorylated IB␤ (lanes 2 -5).
Together, these results suggest that the serine/threonine phosphatase inhibitor calyculin A is able to induce the proteolysis of IB␤ in both Jurkat T cells and 70Z/3 pre-B cells and that the degradation of IB␤ is preceded by its hyperphosphorylation.

TNF-␣ Promotes Calyculin A-induced Degradation of IB␤-
Prior studies have demonstrated that the TNF-␣-elicited cellular activation signal is insufficient to induce the degradation of IB␤, although this signal induces the degradation of IB␣ (12). To investigate whether the TNF-␣-mediated signal could synergize with the phosphatase inhibitor in the degradation of IB␤, we examined the effect of TNF-␣ on calyculin A-mediated degradation of IB␤. As previously reported, incubation of Jurkat T cells with TNF-␣ alone was inefficient in the induction of IB␤ degradation (Fig. 2, lanes 4 and 5). However, when the cells were treated with TNF-␣ together with calyculin A, significant IB␤ degradation could be detected as early as 5 min after cellular stimulation (lane 6), and the entire intracellular pool of IB␤ was almost completely depleted at 30 min poststimulation (lane 7). Consistent with the results shown in Fig.  1A, calyculin A alone induced the degradation of IB␤ (lanes 2 and 3); however, the kinetics of IB␤ degradation in these cells was slower compared to that detected in cells costimulated with TNF-␣ and calyculin A (compare lanes 2 and 3 with lanes  6 and 7).
Both the N-and C-terminal Sequences Are Required for Degradation of IB␤-To further explore the biochemical mechanism underlying the induction of IB␤ degradation by calyculin A and TNF-␣, studies were performed to examine the sequences required for the inducible degradation of IB␤. For these studies, cDNA expression vectors encoding HA-tagged wild type or mutant IB␤ were transfected into Jurkat-Tag cells, and the inducible degradation of these IB␤ proteins was analyzed by immunoblotting using an anti-HA antibody. The exogenously transfected wild type IB␤ expressed as two forms with slightly different mobilities on SDS-PAGE (Fig. 3A, lane  2). As seen in 70Z/3 cells, the differential mobility of these two forms of IB␤ appeared to be due to different levels of phosphorylation as demonstrated by in vitro calf intestinal alkaline phsophatase assays (data not shown). Stimulation of the transfectants with calyculin A and TNF-␣ led to the gradual depletion of the ectopic IB␤ (lanes 3-5). Thus, as seen with its endogenous counterpart, the transfected HA-tagged IB␤ could be degraded in response to cellular stimulation. Deletion of an N-terminal region (amino acids 5-27) covering two potential phosphorylation sites (serines 19 and 23) did not influence the basal phosphorylation of IB␤, since both the slow and fast migrating bands were detected in cells transfected with this mutant (Fig. 3B, lane 4). However, this IB␤ deletion mutant failed to be degraded following cellular stimulation with calyculin A and TNF-␣ (lanes 4 -6). To examine whether serines 19 and 23 were important for the degradation of IB␤, an IB␤ mutant bearing mutations at these sites was tested in the degradation assay. As expected, mutation of these two serines to alanines significantly inhibited the degradation of IB␤ (lanes 1-3). We then examined the potential role of the Cterminal sequences in the degradation of IB␤. In this regard, the C-terminal portion of IB␤ is rich in serines and aspartic acids (12) and has recently been shown to contain the sites for constitutive phosphorylation (41). Consistent with this recent study, an IB␤ mutant lacking the C-terminal 51 amino acids {IB␤(1-308)} migrated on SDS-PAGE as a single band (lane 7), indicating the lack of constitutive phosphorylation. More importantly, deletion of the C-terminal acidic sequences markedly inhibited the degradation of IB␤ (lanes 8 and 9). Thus, degradation of IB␤ induced by calyculin A and TNF-␣ requires both the N-terminal potential phosphorylation sites and the C-terminal sequences. DISCUSSION The nuclear expression and biological function of the NF-B transcription factor is tightly regulated through its cytoplasmic retention by ankyrin-rich inhibitors, including IB␣ and IB␤ (12)(13)(14)(15)(16). Activation of NF-B by various cellular stimuli involves the site-specific phosphorylation and subsequent proteolytic degradation of IB␣, which is associated with the transient nuclear expression of the liberated NF-B factors (18 -21). However, activation of the IB␤-sequestered NF-B pool is triggered by only certain cellular stimuli, which normally induce persistent NF-B activation, such as lipopolysaccharide and interleukin-1 (12) and the HTLV-I Tax protein (32, 33). It remains elusive why the two types of IB molecules differentially respond to the cellular activation signals. Although IB␤ contains two N-terminal serines (Ser-19/Ser-23) that are homologous to the inducible phosphorylation sites of IB␣, it is not clear whether these sites are phosphorylated in response to cellular stimulation.
In the present study, we have demonstrated that the serine/ threonine phosphatase inhibitor calyculin A is able to induce the phosphorylation and degradation of IB␤. This finding supports a model that phosphorylation may trigger the proteolysis of IB␤. However, from our current study, we cannot conclude that the two N-terminal serines (serines 19 and 23) are phosphorylated. To directly address this question, phosphopeptide analyses are necessary. It is clear, though, that site-directed mutagenesis of these two serines to alanines markedly attenuates degradation by calyculin A (Fig. 3) and  1, 4, and 7, respectively) or treated with calyculin A and TNF-␣ for the indicated times. Extract preparation and immunoblotting was performed as mentioned in A.
We have also demonstrated that in Jurkat T cells, TNF-␣ in synergy with calyculin A has the capacity to accelerate the degradation of IB␤. However, in agreement with a previous study (12), in these leukemic T cells, TNF-␣ alone is not sufficient to induce the degradation of IB␤ (Fig. 2), although it is efficient in the induction of IB␣ degradation (27). These results suggest that the signals required for triggering the degradation of IB␤ are more stringent than those for the degradation of IB␣. One possibility is that the N-terminal region of IB␤ is not as efficient a substrate for phosphorylation as the homologous region of IB␣. Phosphorylation of IB␤ thus would require more potent signals which presumably would result in more vigorous kinase activity. Phosphatase inhibitors would act in this manner by leaving kinase activity virtually unopposed by any regulatory cellular phosphatases.
As seen with IB␣, the C-terminal region of IB␤ is rich in prolines, serines, and acidic amino acids. Such a sequence, known as PEST, has been proposed to be involved in the rapid turnover of certain proteins (43). The PEST sequence appears to be dispensable for the inducible degradation of IB␣, although this C-terminal region may regulate the constitutive turnover of IB␣ (22,44,45). However, the PEST sequence is likely required for the inducible degradation of IB␤ since deletion of this region renders IB␤ nonresponsive to the potent degradation signals initiated by calyculin A together with TNF-␣ (Fig. 3B). A recent study (41) suggests that phosphorylation of two serines within the PEST region is critical for the interaction of IB␤ with the c-Rel protooncoprotein. However, it is unclear whether the basal phosphorylation at the C terminus of IB␤ plays a role in regulation of its inducible degradation, although a C-terminal truncation mutant of IB␤ lacking these phosphorylation sites becomes unresponsive to the degradation signals (Fig. 3B). Studies are in progress to determine possible functions of the IB␤ C-terminal PEST domain and to map specific amino acids in the C terminus required for inducible degradation.