PP2A Dephosphorylates Elf-1 and Determines the Expression of CD3ζ and FcRγ in Human Systemic Lupus Erythematosus T Cells¹

T cells from patients with systemic lupus erythematosus (SLE)³ display various biochemical abnormalities, which are invariably associated with altered cell signaling and gene transcription processes and cell function (1, 2, 3, 4). Prominent among these abnormalities is the decreased expression of CD3ζ both at the protein and mRNA levels (5). Various mechanisms appear to contribute to the decreased expression of CD3ζ including decreased CD3ζ gene transcription linked to decreased formation of the DNA-binding form of the transcription factor Elf-1 (6), decreased stability of the CD3ζ mRNA (7, 8), and increased degradation by caspase-3 (9).

Although CD3ζ is decreased in SLE T cells, the FcRγ-chain is up-regulated and appears to populate the CD3/TCR complex (10) and lipid rafts (11). FcRγ, unlike CD3ζ that associates with ZAP70, associates and signals through pSyk. It appears that the FcRγ-pSyk pathway, along with the aggregated rafts in the surface membrane of SLE T cells (11, 12), contribute to the enhanced early CD3/TCR-mediated signaling events (1). The underlying cause for the “exchange” of two signaling chains in the CD3/TCR complex of SLE T cells is unknown. Because CD3ζ and FcRγ protein changes are reflected by similar changes in mRNA levels (5, 10), we considered a link at the transcriptional level, which could explain this observation.

Elf-1 is a 619 aa protein sequence and belongs to the Ets family of transcription factors (13). Elf-1 resides in the cytoplasm as an 80 kDa protein and in the nucleus as a 98 kDa DNA-binding form (14, 15). Elf-1 binds the CD3ζ promoter at two GGAA elements and in their absence the promoter activity is reduced significantly (14, 15). Analysis of the promoter of the FcRγ gene also revealed GGAA motifs in the proximal region, suggesting that Elf-1 may be involved in the regulation of the FcRγ gene. Indeed, Elf-1 suppresses the expression of FcRγ gene in basophil and human T cells (16). Elf-1 has been also reported to suppress the transcription of other genes including FcεRα (17). It appears, therefore, that Elf-1 is a transcription factor that may act as enhancer or repressor. Accordingly, we considered that changes in the levels of expression of the DNA-binding 98 kDa Elf-1 form could account for antithetic changes in the levels of CD3ζ and FcRγ expression.

Posttranslational modifications contribute to the difference in the molecular mass between the 80 and 98 kDa forms of Elf-1. Phosphorylation and O-linked glycosylation contribute to the conversion of 80 to 98 kDa form (14), whereas SLE T cells fail to produce the 98 kDa Elf-1 (6). The decreased expression of 98 kDa Elf-1 in SLE T cells could be the result of either decreased phosphorylation or increased dephosphorylation. Recently, we reported that T cells from patients with SLE, but not other rheumatic diseases, express increased levels and activity of the PP2A (18), which contributes to decreased IL-2 production. We report here that PP2A, which is increased in SLE T cells (18), dephosphorylates Elf-1, promotes the expression of the non-DNA binding 80 kDa Elf-1, and leads to decreased expression of CD3ζ and increased expression of FcRγ.

A total of 24 SLE patients who fulfilled at least 4 of the 11 revised criteria of the American College of Rheumatology for the classification of SLE (19) were enrolled in this study. The patients were 83% African-American, 17% white, with a mean age of 42 ± 12 years and overwhelmingly female (92%). A total of 57% of patients were taking oral prednisone at the time of the study that was temporarily discontinued for least 12 h before the blood was collected. Other immuno-modulatory medications the patients were taking were: hydroxychloroquine (79%), azathioprine (32%), methotrexate (4%), and mycophenalate mofetil (18%). Wherever appropriate, an age-, ethnicity-, and sex-matched healthy volunteer was used as control. Both SLE and healthy volunteers donated 50 ml of blood in heparin-lithium tubes. The purification of T lymphocytes has been described before (20). All T cell cultures, unless stated, were done in RPMI 1640 supplemented with 10% FCS, 50 U/ml penicillin, and 50 μg/ml streptomycin. COS-7 cells were obtained from ATCC and cultured in RPMI 1640 with 10% FCS.

The expression plasmids of the wild-type as well as the mutant PP2A were gifts from Dr. B. A. Hemmings (Friedrich Miescher Institute, Basel, Switzerland). The Elf-1 expression plasmid was a gift from Dr. J. Leiden (University of Chicago, Chicago, IL). Cycloheximide (CHX) was purchased form Sigma-Aldrich. The protein kinase C (PKC) inhibitor Bis was purchased from Calbiochem. Okadaic acid was purchased from Upstate Cell Signaling Solutions. Both PP1 and PP2A as well as Fas Abs were purchased from Millipore. Abs against CD3ζ-chain, Elf-1, Fli-1, Ets-2, and protein kinase N (PKN) are purchased from Santa Cruz Biotechonology; antiactin Ab was purchased from Sigma-Aldrich. The production of Ab against has been described (1). The anti-Fas Ab was purchased from Millipore (clone CH11).

To purify nuclear proteins, cell pellet was first re-suspended in protease inhibitors (10 μg/ml aprotinin, 10 μg/ml leupeptin, and AEBSF) freshly added buffer A (20 mM HEPES, 10 mM KCl, and 0.1 mM EDTA) and incubated on ice for 20 min followed by the addition of 0.6% Nonidet P-40 and subjected to centrifugation. The supernatant was collected and used as cytoplasmic lysate while the pellet was further incubated in protease inhibitors freshly added buffer C (20 mM HEPES, 0.4 M NaCl, and 2 mM EDTA) and vortexed in the cold room for 15 min before they were centrifuged at 13,000 rpm for 15 min. The supernatant was then collected and used as nuclear proteins. For total protein lysates, cells were resuspended in protease inhibitor freshly added RIPA buffer and incubated on ice for 30 min before they were centrifuged at 13,000 rpm for 10 min. The supernatant was then collected and stored in −80°C until use.

The transfection of primary T cells has been reported before (21). Specifically, freshly purified primary T cells were incubated with plasmids in the nucleofector reagent and subjected to electroporation following the manufacturer’s protocol (Amaxa). To transfect COS-7 cells, cells were plated 1 day before. Plasmids were then incubated with Lipofectamine-2000 reagent (Invitrogen) for 20 min before they were dispensed to the cells.

The sequence of the primers used for the PCR amplification are as follows: GCGAGGGGGCAGGGCCTGCATGTGAAG-3′, 5′-AGCCTCTGCCTCCCAGCCTCTTTCTGAG-3′ (CD3ζ-chain); 5′-CGGGGATGAAACAATTGAAACTAT-3′, 5′-CTCGCTGGGTCCACTNTGATGTA-3′. (Elf-1); and 5′-CATGGGTCAGAAGGATTCCT-3′, 5′-AGCTGGTAGCTCTTCTCCA-3′ (actin).

The performance of the EMSA has been described before (6). The sequences of the oligonucleotides used for EMSA are as follows: 5′-TCGAGAACCTCCAGGGCTTCCTGCCTGTGAACCA-3′ (CD3ζ); 5′-CCGCCCGGTCTCTTGTGCAGGAAGGGGAAGGGGCCAAA-3′ (FcRγ).

Elf-1 was purified by affinity chromatography. In brief, nuclear proteins from Jurkat cells were incubated with concatenated oligonucleotides defining the high-affinity binding sites for Elf-1 (14, 22). The identity of the purified proteins was confirmed by Western blotting. The SDS-PAGE purified Elf-1 was subjected to in-gel trypsin digestion. The resulting peptides were spotted onto the MALDI-TOF MS sample plate with matrix (α-cyano-4-hydroxycinnamic acid) and internal mass standards. MALDI-TOF MS measurements were then conducted using a delayed extraction time-of-flight mass spectrometer (Voyager-DE STR Biospectrometry Workstation; Applied Biosystems). After spectrum acquisition, the data file was processed using the instrumental Explore program until a peak list was obtained. The phosphorylated peptides were then identified by comparing the measured masses (peak list) with the calculated peptide masses from theoretical trypsin-digested phosphorylated Elf-1 protein.

Purified T cells were treated with 50 nM okadaic acid or DMSO for 6 h. The cells were then spun down and resuspended in fresh RPMI 1640 medium supplemented with 1% FCS at a concentration of 5 million cells per ml; Indo-1 was added (final concentration 5 μg/ml) in the solution and the cells were incubated at 37°C in the dark for 1 h. The cells were then spun down and resuspended in warm RPMI 1640 plus 1% FCS. Additionally, a CaCl₂ solution was prepared by adding CaCl₂ in RPMI 1640 plus 1% FCS (final CaCl₂ concentration was 1 mM). For each calcium flux experiment, 200 ml of cell solution (1 million cells) were mixed with 800 ml of the CaCl₂ solution. The calcium flux was measured using the LSRII flow cytometer (BD Biosciences) as the ratio of violet to blue emission. The cells were initially run through the machine for 1 min to establish a baseline and thereafter anti-CD3 Ab (final concentration 10 ng/ml) was added followed 30 s later by goat anti-mouse cross-linking Ab. The results were analyzed using the FlowJo software package.

Because the formation of the 98-kDa form of Elf-1 involves posttranslational phosphorylation of the 80-kDa form (14) and SLE T cells express increased amounts of PP2A (18), we asked whether overexpression of PP2A can convert the 98-kDa Elf-1 form to the 80 kDa. Overexpression of wild-type PP2A in COS-7 cells converted Elf-1 from the 98- to the 80-kDa form (Fig. 1,a); overexpression of two mutated PP2A constructs, one affecting His-118, which is crucial for accelerating phosphatase activity and the other affecting Leu-199, which mediates the metal binding and the contact between PP2A and substrate, failed convert to the 98- to the 80-kDa Elf-1 (Fig. 1,a). Wild-type and mutant constructs expressed similar amounts of protein (Fig. 1,b), and therefore, the observed effect on Elf-1 can be attributed to phosphatase activity. Subsequently, we transfected human peripheral blood T cells with wild-type PP2A and we noted that the presence of PP2A decreased the levels of 98-kDa whereas it increased the levels of the 80-kDa Elf-1 (Fig. 1, c and d). These results clearly demonstrate that the 98-kDa Elf-1 is subject to dephosphorylation by PP2A.

PP2A has been shown to trigger apoptosis (23), and SLE T cells have been shown to undergo enhanced spontaneous apoptosis (24). The execution of apoptosis involves extensive activation of proteases that digest protein substrates. We asked whether the increased expression of the 80-kDa and the decreased 98-kDa Elf-1 in SLE T cells could be attributed to enhanced apoptosis. PKN protein (120 kDa, a fatty acid- and ρ-activated Ser/Thr kinase) represents such a protein that during apoptosis is proteolytically digested to generate 105-, 90-, or 55-kDa products, designated apoptotic fragment (AF)1, AF2, and AF3, respectively (25). As shown in Fig. 2, treatment of normal human T cells with anti-Fas Ab (150 ng/ml) resulted in a time-dependent decrease of the 120-kDa PKN and increase of two of its specific proteolytic products, AF1 and AF3. However, this treatment led to increase rather than decrease of the 98-kDa Elf-1, indicating that the conversion of Elf-1 from the 98- to the 80- kD a form is not an apoptosis-related process.

We next asked which of the 60 Ser/Thr residues defined by Elf-1 is responsible for the phosphorylation-mediated conversion of the 80- to 98-kDa Elf-1. Affinity chromatography purified Elf-1 from Jurkat cells was subjected to MS analysis, which demonstrated that Elf-1-derived peptides 230–235 displayed molecular masses equal to those predicted for the corresponding peptides modified by phosphorylation (Table I). Since amino acid Thr-231 is the only amino acid within the peptide 230–235 (WTQREK) that can be phosphorylated, we concluded that Thr-231 represents the residue undergoing phosphorylation during the conversion of the 80- to the 98-kDa Elf-1. Interestingly, the sequence stretching between amino acids 228–234 represents a consensus PKC phosphorylation site. We mutagenized Thr-231 to Val and the resulting plasmid, T231V Elf-1 construct, was transfected into primary T cells. As shown in Fig. 3, a and b (cumulative data), T231V Elf-1-transfected T cells display less 98-kDa and increased amounts of 80-kDa Elf-1. To ascertain that the difference in the expression of 98-kDa Elf-1 in the wild-type and T231V Elf-1-transfected T cells was not due to the defects in used plasmids, the quality and quantity of both plasmids were constantly monitored by sequencing (not shown), Elf-1 insertion sites enzyme restriction (Fig. 3 c), and DNA concentration measurement. Taken together, the data indicate that the conversion of Elf-1 from 80 to 98 kDa involves Thr-231 phosphorylation.

Elf-1 regulates the transcription of genes whose promoters define GGAA motifs. CD3ζ promoter defines such motifs (15), and we found that the FcRγ gene promoter also contains multiple copies of GGAA motif within its −200 bp proximal region (16). An oligonucleotide defined by the -121/-84 region of the FcRγ promoter bound less (p < 0.05) nuclear protein from SLE T cells than from normal T cells (Fig. 4, a and b, cumulative data, n = 4). The presence of an anti-Elf-1 Ab in a shift assay disrupted the binding of normal nuclear protein to the -121/-84 oligonucleotide (16) indicating the specific binding of Elf-1. Overexpression of PP2A in fresh human T cells resulted in decreased expression of 98 kDa Elf-1, but not Ets-2, in the nucleus (not shown) and nuclear proteins from PP2A-transfected cells displayed limited binding to CD3ζ and FcRγ promoter-defined oligonucleotides (Fig. 4 c). These data indicate that PP2A can regulate the gene transcription of both CD3ζ and FcRγ genes by decreasing the binding of Elf-1 to their promoters.

Next, we tested whether silencing of the overexpressed PP2A in SLE T cells by small interfering RNA (siRNA) PP2A can restore the expression of CD3ζ and suppress the increased expression of FcRγ in SLE T cells. To silence PP2A, we used a mixture of three PP2A-specific siRNA (PP2A siRNA-mix) as well as each of the three PP2A siRNA constructs separately (PP2A siRNA-1, PP2A siRNA-2, and PP2A siRNA-3). Successful silencing of PP2A was confirmed by RT-PCR (Fig. 5,a) and real-time RT-PCR (not shown). PP2A mRNA levels were not affected when the cells were treated with scrambled siRNA or siRNA-targeting SRP40 mRNA. As shown in Fig. 5, a and b (cumulative data), CD3ζ mRNA was only marginally increased despite of the significantly decreased PP2A mRNA in the PP2A siRNA-mix transfected cells. In contrast, this treatment was sufficient to decrease the expression of FcRγ mRNA (Fig. 5, a and b). Because CD3ζ-chain mRNA in SLE T cells is subject to enhanced degradation (7), which could offset the siRNA PP2A-induced increase in CD3ζ mRNA, we treated the siRNA PP2A-mix transfected SLE T cells with CHX, an inhibitor of protein synthesis commonly used to block new protein synthesis mediating the degradation of mRNA. As shown in Fig. 5, c and d (cumulative data), the presence of CHX increased significantly the expression of CD3ζ mRNA. To address whether the effect of PP2A siRNA-mix on FcRγ mRNA levels is specific, we transfected T cells with each individual siRNA PP2A duplex separately. As shown in Fig. 5 e, these constructs decreased the expression of PP2A but not those of GAPDH and CD3ε. The moderate decrease in PP2A expression was associated with a decrease in FcRγ mRNA levels albeit to a lesser degree than the PP2A siRNA-mix, establishing further the specificity of this approach. Taken together, our findings indicate that increased expression of PP2A is at least in part responsible for the aberrant expression pattern of both CD3ζ and FcRγ in SLE T cells, and suppression of the expression of PP2A can restore normal expression pattern for both CD3ζ- and FcRγ-chains in SLE T cells.

To determine whether inhibition of PP2A activity in SLE T cells with a pharmacologic agent would correct established aberrant cell function, we treated cells with okadaic acid (50 nM) for 6 h and the cell lysates were blotted with anti-CD3ζ, FcRγ, and actin Abs. As seen in Fig. 6,a, cells treated with okadaic acid expressed more CD3ζ-chain and less FcRγ-chain compared with cells treated with DMSO. Because expression of FcRγ in SLE T cells contributes to increased CD3-mediated free intracytoplasmic calcium response (1), we treated SLE T cells with okadaic acid or DMSO for 6 h and recorded free intracytoplasmic calcium response as the violet/blue emission of Indo-1 loaded cells. The representative tracing (four different patients were studied with similar tracings) of cells after treatment with anti-CD3 Ab, followed by a cross-linking Ab, shows that okadaic acid corrects the increased free intracytoplasmic calcium response (Fig. 6 b). Treatment of SLE T cells with okadaic acid for 18 h yielded the same effect on the calcium response. Similarly treated normal cells with the same dose of okadaic acid for 6 or 18 h did not affect the calcium response (not shown).

The findings in the current study demonstrate that Elf-1 can function as a molecular valve that regulates the expression of CD3ζ- and FcRγ-chains in an opposing manner. This finding echoes previous findings that Elf-1 can act either as a transcription activator or suppressor (26). Although it is not clear at the present time how Elf-1 can serve as a transcriptional enhancer or repressor, it has been shown to be part of a multi-molecule complex that regulates the expression of many genes active in the hemopoietic system (27). Whether Elf-1 is an activator or suppressor probably depends on the molecular partners it interacts with. Comparison of the CD3ζ and FcRγ promoters revealed that although they both contain GGAA sites and bind Elf-1, the flanking sequences are not the same, allowing the binding of different transcription factors that may provide proper context for Elf-1 to behave as an enhancer or repressor. In addition, Elf-1 encodes more than 60 amino acids that can either be phosphorylated or O-linked glycosylated. We previously showed that the 98 kDa form exists in three forms that can be distinctly separated by both SDS and isoelectric focusing gels (6). It is possible that each of these three forms has distinct ability to interact with other factors involved in the regulation of transcription.

Increased PP2A expression in SLE T cells was previously shown to dephosphorylate phosphorylated CREB and limit is binding to the c-fos and IL-2 promoters leading to decreased expression of AP1 and IL-2 (18). The current study demonstrates that PP2A can dephosphorylate Elf-1 and alter the transcription CD3ζ and FcRγ genes, the products of which are components of the CD3/TCR complex in SLE T cells (1). We show that modulation of the levels of PP2A in T cells either by overexpression or silencing results in changes of the expression of CD3ζ and FcRγ mRNA in an opposite manner. We also demonstrate that inhibition of PP2A with okadaic acid suppresses the expression of FcRγ while it increases the expression of CD3ζ-chain. Therefore, the PP2A/Elf-1 pathway determines to a point the composition of CD3 complex. Low levels of PP2A result in CD3ζ expression, whereas increased expression of PP2A, as is the case in SLE T cells, results in decreased levels of 98 kDa Elf-1, which limits the expression of CD3ζ and allows that expression of FcRγ, which becomes a component of the CD3/TCR complex. The effect of PP2A is counterbalanced by PKC but, in SLE T cells, it probably remains unopposed because the expression and activity of PKC in SLE T cells has been reported decreased (18, 28).

We have observed that overexpression in primary T cells of Elf-1 T231V, as compared with the wild-type Elf-1, resulted in the decreased expression of the 98 kDa and simultaneously increased expression of the 80 kDa Elf-1. This data indicates that T231 is crucial for the conversion of Elf-1 from 80 to the 98 kDa. It is not clear though whether phosphorylation of T231 is sufficient for this conversion or phosphorylation at other sites and/or other posttranslational modifications are also involved. Phosphorylation has been shown to be coupled with other posttranslational modifications such as O-linked glycosylation (29), ubiquitination (30). Furthermore, proteins are commonly phosphorylated at multiple sites following a hierarchical order (31). It is possible that the dephosphorylation of T231 prevents other amino acids from being phosphorylated or modified by other posttranslational modalities and altogether these defects account for the decreased conversion of Elf-1 from 80 to 98 kDa.

We have shown that the conversion from the 98 to the 80 kDa form of Elf-1 is not the result of a Fas-related apoptotic event despite the fact that PP2A has been shown to trigger the onset of apoptosis (23) and SLE T cells are known to display enhanced spontaneous apoptosis (24). Instead, our current data and those reported earlier (6, 14) have established that this conversion is a phosphorylation-dependent event.

We found that additional treatment with CHX, an inhibitor of RNA destabilizing factors, is needed for the siPP2A-transfected SLE T cells to stabilize CD3ζ mRNA. This finding is in agreement with the well-established observation that CD3 ζ mRNA is unstable in SLE T cells (7, 8) and it appears that CHX-sensitive proteins are involved. Previous reports have shown that CHX increases the mRNA stability of genes including IL-6 (32), c-fos (33), and SOX-9 (34) through a mechanism known as superinduction.

Our findings establish that PP2A binds to and dephosphorylates Elf-1 at Thr-231 and that modulation of the levels of PP2A either by forced expression or silencing results in altered CD3ζ and FcRγ promoter activity. We propose that the PP2A>Elf-1>CD3ζ/FcRγ pathway is of particular significance in SLE T cells because the established increased PP2A activity in SLE T cells (18) results in a structurally variant CD3 complex that contains and handles TCR-initiated signaling not through the normally used ζ-chain but rather through the FcRγ-chain. We previously demonstrated that silencing of PP2A in SLE T cells results in corrected production of IL-2 by limiting the dephosphorylation of the IL-2 transcriptional enhancer phosphorylated CREB (18). Our current data prove that enhanced IL-2 production after reduction of PP2A may come through corrected CD3ζ-chain expression since replenishment of CD3ζ in SLE T cells also results in increased production of IL-2 (35).

The authors have no financial conflict of interest.