Systemic lupus erythematosus (SLE) is an autoimmune disorder of indeterminate etiology characterized by abnormal T cell signal transduction and altered T cell effector functions. We have previously observed a profound deficiency of total protein kinase A (PKA) phosphotransferase activity in SLE T cells. Here we examined whether reduced total PKA activity in SLE T cells is in part the result of deficient type II PKA (PKA-II) isozyme activity. The mean PKA-II activity in SLE T cells was 61% of normal control T cells. The prevalence of deficient PKA-II activity in 35 SLE subjects was 37%. Deficient isozyme activity was persistent over time and was unrelated to SLE disease activity. Reduced PKA-II activity was associated with spontaneous dissociation of the cytosolic RIIβ2C2 holoenzyme and translocation of the regulatory (RIIβ) subunit from the cytosol to the nucleus. Confocal immunofluorescence microscopy revealed that the RIIβ subunit was present in ∼60% of SLE T cell nuclei compared with only 2–3% of normal and disease controls. Quantification of nuclear RIIβ subunit protein content by immunoprecipitation and immunoblotting demonstrated a 54% increase over normal T cell nuclei. Moreover, the RIIβ subunit was retained in SLE T cell nuclei, failed to relocate to the cytosol, and was associated with a persistent deficiency of PKA-II activity. In conclusion, we describe a novel mechanism of deficient PKA-II isozyme activity due to aberrant nuclear translocation of the RIIβ subunit and its retention in the nucleus in SLE T cells. Deficient PKA-II activity may contribute to impaired signaling in SLE T cells.
Systemic lupus erythematosus (SLE)3 is an autoimmune disorder of indeterminate etiology characterized by impaired cellular immunity (1) that often results in anergy to recall Ags (2, 3). Within the T cell compartment, both the CD3, CD4 helper and CD3, CD8 cytotoxic/suppressor subsets are dysfunctional, resulting in an imbalance of exaggerated helper and diminished cytotoxic/suppressor activities (4). A marker of altered CD4 helper function is skewing of the Th1 and Th2 responses (5) to antigenic challenge. There is diminished Th1 and enhanced Th2 cytokine production, yielding reduced generation of IL-2 and IFN-γ by Th1 cells and overproduction of IL-6 and IL-10 by Th2 cells (6, 7, 8, 9). However, the mechanisms contributing to T cell immune dysfunctions in SLE are still incompletely understood (10).
One mechanism that may contribute to T cell dysfunction in SLE is altered signal transduction. We have identified the presence of several, discrete signaling defects in SLE T cells (10, 11). Deficient type I protein kinase A (PKA-I) activity is a signaling disorder that results in marked underphosphorylation of substrates (12, 13) and occurs with a prevalence of 80% in SLE T cells (14). Kinetic analyses of this isozyme deficiency revealed a significant reduction in both the maximal enzyme velocity (Vmax) and cAMP-binding capacity of the type I regulatory (RI) subunit (Bmax), but a significant increase in the cAMP half-maximal activation (Ka) of the PKA-I holoenzyme compared with controls (15). This altered isozyme kinetics is the result of a significant reduction of both RIα and RIβ isoforms in SLE T cells (16). The association of diminished RI protein content with reduced amounts of steady state RI transcripts raises the possibility that the PKA-I isozyme deficiency could reflect a pretranslational disorder in SLE T cells. By hindering efficient phosphorylation of multiple substrates, deficient PKA-I activity would be expected to significantly impede signaling downstream. Because PKA-catalyzed phosphorylation is a principal posttranslational process that regulates widely divergent cellular functions, including chaperonin activities (17), binding of agonists to intracellular receptors (18, 19), catalysis (20, 21), and activation of transcription factors (22, 23, 24), deficient PKA-I activity may contribute to altered helper activity and cytotoxicity in SLE T cells.
PKA is a serine/threonine kinase that is composed of two isozymes, type I PKA (PKA-I) and type II PKA (PKA-II) (25). These isozymes differ in their subcellular localization; in the human T cell, PKA-I localizes predominantly with the plasma membrane fraction whereas PKA-II is present chiefly in the cytosol (26). The R subunits, RI and RII, are comprised of highly homologous α and β isoforms (i.e., RIα/β and RIIα/β) and the catalytic (C) subunits of α, β, and γ isoforms. In their holoenzyme forms, both isozymes exist primarily as homodimers, RIα/β2C2 and RIIα/β2C2. In the T cell, PKA isozymes can be activated via two mechanisms. Occupancy by an agonist of Gs-bound stimulatory receptors (Rs) activates adenylyl cyclase (AC), hydrolyzing ATP to cAMP. Alternatively, binding of antigenic peptide to the TCR initiates a signal from the CD3 complex that bifurcates at the level of protein kinase C, phosphorylates AC, and stimulates AC catalysis and cAMP turnover (27). Binding of cAMP to the A- and B-binding domains of the R subunits activates the holoenzymes, as shown by the equation: R2C2 + 4cAMP ↔ R2cAMP4 + 2C (25).
The present study was undertaken to explore the idea that the profound deficiency of total PKA activity in SLE T cells is in part the result of deficient PKA-II phosphotransferase isozyme activity. During the course of our work, we have observed that markedly diminished total PKA activity in SLE T cells was associated with a concomitant reduction of PKA-II activity in the cells of some subjects with deficient PKA-I isozyme activity (13). To document PKA-II isozyme deficiency, we performed a prospective analysis of 35 unselected, consecutive SLE subjects and controls. This analysis revealed that: 1) SLE T cells can harbor a significant, co-existent reduction of PKA-II activity; 2) the prevalence of deficient PKA-II activity in this cohort is 37%; 3) like deficient PKA-I activity, there is no apparent relationship of deficient PKA-II activity and SLE disease activity; and 4) the mechanism of this isozyme deficiency is aberrant translocation of the RIIβ isoform to the nucleus from the cytosol and its retention in the nucleus. This association of nuclear translocation of a protein kinase regulatory subunit with a persistent deficiency of protein kinase activity is a novel mechanism in primary T cells. Moreover, this mechanism is distinct from that of deficient PKA-I isozyme activity (16). Together, deficient PKA-I and PKA-II activities yield a profound reduction of total cAMP-activatable PKA activity, which may significantly impede TCR-initiated signaling (27) and contribute to compromised T cell effector functions in SLE (10).
Materials and Methods
Patient and control populations
Thirty-five consecutive, unselected SLE subjects with a mean age (±SD) of 35.5 ± 11 years (range, 12–66 years) were prospectively studied. All subjects fulfilled four or more of the criteria for the classification of SLE (28). Of these SLE subjects, 29 were female, 26 were white, and 9 were black. Utilizing the SLE disease activity index (SLEDAI), a standardized scale, to gauge the extent of disease manifestations (29), the mean (±SD) SLEDAI was 11.9 ± 7.4 (range, 2–32). A SLEDAI of 1–10 denotes mild activity, 11–20 moderate activity, and ≥21 severe disease activity (13, 14, 15, 16). Thirty-five healthy controls with a mean age of 35.1 ± 9.5 years (range, 24–65 years) were studied. Of these controls, 23 were female, 24 were white, and 11 were black. Eleven subjects with primary Sjögren’s syndrome (SS) (30) with a mean age of 44.2 ± 5.0 years (range, 28–56 years) served as disease controls. Of these, all subjects were white and female.
Subjects were studied according to our previous protocols (12, 13, 15, 16). Individuals experiencing a flare of SLE activity were studied before initiation of corticosteroid and/or immunosuppressive therapy; none had been treated with immunosuppressive agents for at least 3 mo. Only SLE subjects treated with low dose corticosteroids (≤10 mg/day prednisone) were entered into this protocol; these individuals were studied 24 h after their last oral dose. Nonsteroidal antiinflammatory agents and hydroxychloroquine were withheld for 72 h and 7 days, respectively, before study when clinically feasible. Informed consent to participate in this study and to obtain specimens by venipuncture or by leukapheresis were obtained from subjects and controls. The research protocols and consent forms were approved by the institutional review board of the Wake Forest University/Baptist Medical Center.
T lymphocyte isolation and phenotypic characterization
SLE and control T lymphocytes were isolated and enriched from PBMC or cells obtained by leukapheresis by the high gradient magnetic cell separation system, Midi MACS (Miltenyi Biotec, Auburn, CA) (16). Cytofluorographic analysis revealed that enriched, viable T cells expressed a mean (±SEM) of 96 ± 1.2% CD3. The proportions of CD3 T cells expressing other cell surface markers have been previously detailed (14).
T cell lines
T cell PKA-I and PKA-II isozymes were fractionated and isolated by tandem DE52-cellulose and carboxymethyl-Sephadex chromatography as previously described (13). PKA phosphotransferase activity was quantified as previously detailed (27). The physiologic range of T cell PKA-II activity is given in Table I.
|.||PKA-II sp. act. (pmol/min/mg protein) .||.|
|.||Controls (n = 35) .||SLE (n = 35) .|
|Mean± SD||481.2 ± 144.1||293.8 ± 192|
|.||PKA-II sp. act. (pmol/min/mg protein) .||.|
|.||Controls (n = 35) .||SLE (n = 35) .|
|Mean± SD||481.2 ± 144.1||293.8 ± 192|
Nuclear extracts were prepared from the T cells of normal and SS disease controls and SLE subjects. T cells (80 × 106) were washed twice in PBS and resuspended in 1 ml ice-cold lysis buffer (Dignam buffer A: 10 mM Tris-HCl (pH 7.4), 10 mM NaCl, 3 mM MgCl2, 0.1 mM EGTA, 0.5 mM DTT, 1 mM PMSF, and 2 μg/ml each leupeptin and aprotinin). After 10 min on ice, 50 μl 10% Nonidet P-40 were added, and the cells were centrifuged at 9000 rpm for 30 s at 4°C. Pelleted nuclei were washed twice in Dignam buffer A and lysed in 200 μl Dignam high salt buffer C (20 mM HEPES (pH 7.9), 420 mM NaCl, 1.5 mM MgC12, 0.1 M EDTA, 25% glycerol, 0.5 mM DTT, 0.5 mM PMSF, 2 μg/ml each leupeptin and aprotinin) for 15 min at 4°C. After lysis, nuclear extracts were centrifuged at 12,000 rpm for 10 min at 4°C, and the resulting supernatants were diluted 1:1 (v/v) with Dignam buffer D (20 mM HEPES (pH 7.9), 100 mM KCl, 0.1 mM EDTA, 20% glycerol, 0.5 mM DTT, 0.5 mM PMSF, 2 μg/ml each leupeptin and aprotinin). Protein concentration was determined by the Bradford method (31) (Bio-Rad Laboratories, Hercules, CA).
Immunoprecipitation and immunoblotting
Immunoprecipitation and immunoblotting were performed as previously described (16, 27). Nuclear extracts (20 μg) were incubated overnight with 1:250 anti-RIIβ mAb (Transduction Laboratories, Lexington, KY). Immune complexes were isolated by using affinity-purified goat anti-mouse IgG conjugated to protein A-Sepharose. RIIβ was then eluted by boiling the immune complexes in 25 μl buffer A (50 mM Tris-HCl (pH 6.8), 30% (v/v) glycerol, 0.025% bromphenol blue, 2% SDS, and 10% 2-ME) for 3 min at 95°C. Samples were resolved by 10% one-dimensional SDS-PAGE. Immunoblots were prepared, probed with 1:1,000 anti-RIIβ mAb, and developed with enhanced chemiluminescence (ECL) (16).
Nuclei-free T cell homogenate was also prepared as previously described (16). Following separation of T cell homogenate (200 μg) by one-dimensional SDS-PAGE and transfer to Immobilon-P (Millipore, Bedford, MA), membranes were immunoblotted with 1:250 anti-RIIα mAb (Transduction Laboratories) or 1:1000 anti-RIIβ mAb, washed with buffer C (100 mM Tris-HCl (pH 7.5), 500 mM NaCl, and 0.1% Tween 20), and probed with 1:4000 HRP-labeled sheep anti-mouse IgG in Blotto. After four washings with buffer C, the blots were developed by ECL. Primary and secondary Abs were then extracted from the membrane using buffer D (62.5 mM Tris-HCl (pH 6.7), 100 mM 2-ME, and 2% SDS) (16) and were reprobed with 1:100 polyclonal rabbit anti-human actin, 1:4000 HRP-labeled sheep anti-rabbit IgG, and ECL. Quantification of RIIα and RIIβ proteins in total T cell protein or isolated nuclei was performed by laser densitometry, the amounts calculated from reference standard curves, and expressed as arbitrary densitometric units (ADU) (16).
Confocal immunofluorescence microscopy was performed as previously detailed (17). Cells were centrifuged at 600 × g, transferred to Eppendorf tubes, and centrifuged at 4000 rpm for 4 min at 4°C. The pellet was resuspended in 100 μl fixation buffer (4.0% paraformaldehyde, 120 mM sucrose in PBS) per tube, and fixation was allowed to proceed at 4°C for 30 min (32). PBS, 1 ml, with 1 mg/ml BSA (PBS/BSA) was added to each tube and centrifuged, and the pellet was resuspended in 100 μl quench solution (50 mM NH4Cl in PBS) to stop the fixation. After the pellet was washed twice in PBS/BSA, fixed cells were resuspended in 100 μl permeabilization buffer (0.2% Triton X-100, 1 mg/ml BSA in PBS) containing 1:50 anti-RIIα, anti-RIIβ, or anti-Cα subunit mAbs, and the cells were incubated for 60 min at room temperature. After permeabilization of the cells, the resulting pellet was resuspended in permeabilization buffer containing 1:50 FITC-F(ab)2 goat anti-mouse IgG. This suspension was incubated for 45 min at room temperature, the labeled cells were washed twice with permeabilization buffer, and the cells were then incubated in 5 μM PI to label nuclei. After the cells were washed, the pellet was resuspended in 20 μl of the DABCO/Mowiol solution (33, 34) and examined with a Zeiss LSM 510 confocal microscope.
The prevalence of PKA-II isozyme deficiency is defined as the probability of currently having that isozyme deficiency regardless of the duration of time one has had the disorder. It is calculated by dividing the number of subjects with the isozyme deficiency by the number of subjects in the study population. Statistical significance (p = 0.05) was calculated by the paired Student t test, Mann-Whitney U rank-sum test, or ANOVA (SigmaStat, Jandel Scientific, Corte Madera, CA). Except where indicated, means (±SEM) are used throughout the text.
Deficiency of PKA-II isozyme activity in SLE T lymphocytes
To quantify PKA-II isozyme phosphotransferase activity, we isolated PKA-II holoenzyme in T cell homogenates by tandem DE52-cellulose and carboxymethyl-Sephadex column chromatography using a salt gradient. Compared with the PKA-I holoenzyme that elutes between 85 and 110 mM NaCl, the PKA-II holoenzyme characteristically elutes from DE52-cellulose columns between 180 and 220 mM NaCl (26). The mean (±SD) PKA-II phosphotransferase sp. act. in normal T cells from 35 subjects was 481.2 ± 144.1 pmol/min/mg protein. By contrast, T cells from SLE subjects exhibited a mean (±SD) PKA-II activity of 293.8 ± 192 pmol/min/mg (Fig. 1,A and Table I). Although there was some overlap between SLE and controls, the mean PKA-II activity in SLE T cells was 61% of controls (p ≤ 0.001 by paired Student’s t test). If deficient PKA-II isozyme is defined as phosphotransferase activity ≤ 193 pmol/min/mg (i.e., = 2 SD), then the prevalence of deficient PKA-II isozyme activity in this SLE cohort is 37% (13 of 35 subjects).
Relationship of PKA-II isozyme activity to disease activity or therapy
The relationship between PKA-II isozyme activity and SLE disease activity was analyzed to establish whether or not deficient isozyme activity is related to SLE disease activity. There were no significant differences between mean PKA-II activities in subjects with severe, moderate, or mild disease activity. Moreover, the 13 subjects with deficient PKA-II activities were distributed among these groups. Their mean PKA-II activity was 103.8 ± 56.8 pmol/min/mg (25–75%, 72–140 pmol/min/mg; Fig. 1,B, p ≤ 0.001). Unexpectedly, the mean SLEDAI score of SLE subjects with physiologic PKA-II activities was 12.4 compared with 10.6 in subjects with deficient PKA-II activities (Fig. 1 B). Although this difference between the SLEDAI scores was not statistically significant, SLE subjects with physiologic PKA-II activities actually exhibited greater clinical disease activity than those with lower isozyme activities. These data suggest that deficient PKA-II isozyme activity may not necessarily be associated with SLE disease activity.
To determine whether PKA-II activity was associated with therapy, the medical regimen of each subject at enrollment and at three subsequent intervals during 4 years was analyzed. There was no statistical relationship between any medical therapy and PKA-II isozyme activities in SLE subjects with mild, moderate, or severe SLE activity. In five instances, comparison of PKA-II activities before and after corticosteroid therapy demonstrated no significant differences between PKA-II activities. Moreover, five subjects whose disease became clinically inactive and were able to discontinue therapy revealed no significant change in their PKA-II activities over time (data not shown). Together, these results suggest that therapy is unlikely to modify T cell PKA-II activity and, therefore, is also unlikely to be implicated in this T cell isozyme deficiency in SLE.
Persistence of PKA-II isozyme deficiency over time
To determine whether deficient PKA-II activity persists over time and remains independent of disease activity, a group of 15 SLE patients with an initial mean (±SD) SLEDAI score of 12.1 ± 7.2 (25–75%, 6.5–15.5) was followed up to 4 years and restudied on at least three occasions. Of these, six initially had mild SLE activity, eight had moderate activity, and one had severe activity. Fig. 2 demonstrates that there were essentially no differences in PKA-II activities during the follow-up interval. In contrast, subjects being treated for SLE experienced a significant reduction in their SLEDAI scores between the first and second follow-up studies (p = 0.02), but no significant change between the second and third follow-up analyses. Thus, low PKA-II activities persist over time and are independent of disease activity.
In vitro culture does not reverse deficient SLE T cell PKA-II isozyme activity
To determine whether or not deficient SLE T cell PKA-II isozyme activity is reversible, we established T cell lines and studied cells that had been propagated over 10 passages. The advantage of this approach is that it is possible to study the progeny of SLE T cells that were originally isolated from PBMC but have not been exposed to the disease process. Thus, any identified defects cannot be attributed to extracellular stimuli, such as cytokines or immune complexes, that may be present in the lupus microenvironment in vivo.
T cell lines from three untreated SLE patients with markedly reduced PKA-II activities and three healthy controls were established concomitantly. After 10 passages, cycling T cells were harvested, the proportions of cells in each phase of the cell cycle were determined, and PKA-II isozyme activities were quantified. Both SLE and control T cell lines had equivalent proportions of cells in each phase of the cell cycle; 35% of the cells were in S phase. Compared with a mean PKA-II activity of 164.5 ± 53 pmol/min/mg in freshly isolated SLE T cells in G0/G1, cells in S phase had a 29.2% reduction of the mean PKA-II isozyme activity to 116.5 ± 10.6 pmol/min/mg (p = NS). By contrast, the mean PKA-II isozyme activity in control T cell lines in S phase was reduced by a mean 58.6% (224.7 ± 146 pmol/min/mg) compared with freshly isolated T cells in G0/G1 (542.6 ± 187 pmol/min/mg) (p = 0.017). These results reveal that a proportion of the total PKA-II holoenzyme is activated in cycling T cells, resulting in a reduction in the amount of residual PKA-II holoenzyme. A similar effect of T cell proliferation on PKA-I activity has been previously observed (35). That the magnitude of PKA-II activation and utilization in SLE T cells was diminished by one-half (29.2% vs 58.6%) may reflect the low PKA-II activity in G0/G1 cells.
SLE and control T cells were then rested for 72 h in low FCS-containing media in the absence of cytokines and mitogen to force cells to reenter G0/G1 phase of the cell cycle, and PKA-II activity was quantified in these cells. Although >95% of cells had returned to G0/G1 phase, PKA-II activity remained depressed in SLE cells but had increased 30% (298.3 ± 90 pmol/min/mg) toward baseline levels in control cells. Quantification of PKA-II activity in T cells cultured for >72 h in media containing very low concentrations of FCS and no cytokines or mitogens was unreliable due to increasing cell death. Together, these data suggest that the progeny of SLE T cells have a persistent deficiency of PKA-II activity that is independent of cell activation and mitogenesis as well as the lupus microenvironment.
Analysis of RII isoform content in normal and SLE T cell lines
That normal cycling T cells undergo a reduction of PKA-II holoenzyme activity that is partially repleted during the resting phase of the cell cycle raised the possibility that the content of a RII isoform comprising the holoenzyme could be altered. Because the PKA-II isozyme is predominantly localized within the cytosol in human T cells (26), a reduced amount of cytosolic RIIβ and/or RIIα protein is one mechanism that could yield diminished PKA-II phosphotransferase activity. To test this idea, nuclei-free T cell homogenates were prepared from 1) freshly isolated T cells, 2) cycling cells (after 10 passages), and 3) rested T cells (at 72 h). The homogenates were fractionated by 10% one-dimensional SDS-PAGE, electroblotted to Immobilon membrane, and immunoblotted with anti-RIIα and anti-RIIβ mAbs. Fig. 3 demonstrates that freshly isolated control T cells possessed both cytosolic RIIβ and RIIα isoforms. After 10 passages, cycling T cell progeny from controls expressed increased amounts of cytosolic RIIα protein, but no detectable cytosolic RIIβ isoform. After resting cells for 72 h, at which time ≥95% of cells were in G0/G1 phase, control T cells reexpressed cytosolic RIIβ protein, and the amount of RIIα returned toward baseline. Thus, RIIβ was depleted from the cytosol whereas RIIα accumulated in the cytosol of cycling normal T cells. This depletion of RIIβ protein, and therefore RIIβ2C2 holoenzyme, may account for the reduction of PKA-II activity during mitogenesis.
Recognition that reduced PKA-II activity was associated with depletion of RIIβ in normal T cells prompted us to determine whether cytosolic RIIβ is present in freshly isolated SLE T cells. Fig. 3 shows that cytosolic RIIβ, but not RIIα, was absent in freshly isolated SLE T cells. Like cycling normal T cells, cycling SLE T cells revealed no cytosolic RIIβ; instead, there was accumulation of RIIα (Fig. 3). However, in contrast to normal rested T cells, cytosolic RIIβ failed to be reexpressed in rested SLE T cells (Fig. 3) in which ≤4% of cells were in S phase. Moreover, the amount of cytosolic RIIα remained increased. Taken together, these results suggested that a disorder of RIIβ regulation might be associated with deficient PKA-II activity in SLE T cells.
Quantification of T cell RIIα and RIIβ protein content in SLE and controls
To determine whether cytosolic RIIβ protein is reduced or absent in SLE T cells, we examined freshly isolated T cells from 21 SLE subjects, 21 healthy controls, and 11 SS disease controls. The immunoblots shown in Fig. 4 demonstrated 1) the presence of cytosolic RIIβ and RIIα proteins in normal controls in a ratio of 3.95:1, 2) decreased cytosolic RIIα and increased cytosolic RIIβ in SS subjects yielding an increased ratio of 5.35:1, and 3) absence of cytosolic RIIβ and increased cytosolic RIIα protein in SLE subjects. The T cells of all SLE subjects shown in Fig. 4 had deficient PKA-II activity. Table II shows that, on average, there was a 60% reduction of cytosolic RIIβ in SLE T cells, yielding a significantly reduced RIIβ:RIIα ratio of 1.28:1 compared with normal and SS controls. That both normal and SS control T cells have significantly greater cytosolic RIIβ content than SLE T cells suggests disease specificity. In sum, these results demonstrate that deficient PKA-II activity in SLE T cells is associated with reduced cytosolic RIIβ protein content. In 29% (6 of 21) of subjects, there was no detectable cytosolic RIIβ protein.
|RII Isoforms .||Controls .||.||Disease .||p .|
|.||Normal .||SS .||SLE .||.|
|RIIα||1.00 ± 0.13||1.58 ± 0.36||1.24 ± 0.18||0.371,b 0.46,c 0.377d|
|RIIβ||3.95 ± 0.42||8.46 ± 1.01||1.59 ± 0.49||<0.001,b <0.001,c <0.001d|
|RII Isoforms .||Controls .||.||Disease .||p .|
|.||Normal .||SS .||SLE .||.|
|RIIα||1.00 ± 0.13||1.58 ± 0.36||1.24 ± 0.18||0.371,b 0.46,c 0.377d|
|RIIβ||3.95 ± 0.42||8.46 ± 1.01||1.59 ± 0.49||<0.001,b <0.001,c <0.001d|
ADU ± SEM.
Normal vs SS.
Normal vs SLE.
SS vs SLE.
Aberrant nuclear translocation of RIIβ from the cytosol in SLE T cells
Utilizing confocal immunofluorescence microscopy, we have recently demonstrated that activation of the PKA-II isozyme in normal T cells by either 8-chloro-cAMP (8-Cl-cAMP) or anti-CD3 + anti-CD28 + rIL-1α induced nuclear translocation of RIIβ within 30 min that peaked by 1 h (36). Here, we observed that SLE T cells exhibited spontaneous translocation of RIIβ from the cytosol to the nucleus in the absence of in vitro cell activation (Fig. 5,B). On average, 60% of SLE T cells had detectable nuclear RIIβ, whereas only 3% of normal T cells exhibited nuclear translocation of RIIβ by confocal immunofluorescence microscopy (Fig. 6). None of the T cells from normal or SS controls shown in Fig. 5,B had detectable nuclear RIIβ. By contrast, the RIIα isoform remained localized to the cytosol in SLE T cells (Fig. 5,A). Because it has been well demonstrated that the C subunit can diffuse between the cytoplasm and nucleus and can, therefore, be present in both compartments (37), we anticipated the presence of the C subunit in both compartments of T cells (Fig. 5 C).
To quantify the amounts of nuclear RIIβ protein in SLE and controls, we immunoprecipitated RIIβ from nuclear extracts and quantified RIIβ proteins on immunoblots by densitometry. Unexpectedly, RIIβ protein was constitutively present in the nuclei of both normal and SS disease controls (Fig. 5 D). It is likely that RIIβ was identified by immunoblotting, but not by confocal immunofluorescence microscopy, because the amount of constitutive nuclear RIIβ in control T cells was below the limits of detection by confocal immunofluorescence microscopy. By immunoblotting, there was a 54 and 39% increase in nuclear RIIβ protein in SLE T cells compared with normal and SS controls, respectively (n = 6, SLE vs normal controls, p ≤ 0.001). Thus, exaggerated nuclear translocation of RIIβ appears to be the mechanism responsible for deficient PKA-II activity in SLE T cells.
Persistence of nuclear RIIβ after in vitro passage of SLE T cells
Absence of cytosolic RIIβ protein in freshly isolated SLE T cells and its persistent absence in rested T cell progeny raised the question, “What is the disposition of RIIβ?” To determine whether RIIβ might be retained in the nucleus, nuclear extracts from freshly isolated SLE T cells, cells cycled through 10 passages, and cells rested over 72 h were immunoprecipitated, gel separated and electrotransferred, and immunoblotted with anti-RIIβ mAb. Based on quantification by densitometry, there was no appreciable change in the amount of nuclear RIIβ protein in in vitro-rested T cell progeny compared with freshly isolated or in vitro-propagated T cells (data not shown). These results were consistent with the continued absence of cytosolic RIIβ in SLE T cells and suggested that, after its enhanced nuclear translocation, RIIβ is retained in the nucleus of SLE T cells and that this retention is independent of the cell cycle.
Dexamethasone does not alter nuclear RIIβ subunit content in normal T cells
Although we observed no in vivo effect of corticosteroids on PKA-II activity, we considered the possibility that corticosteroids might modify T cell RII subunit protein content or its subcellular localization. To test this possibility, freshly isolated T cells from three healthy controls were cultured in the absence or presence of 10 nM dexamethasone for 18 h, and total T cell lysates were separated by SDS-PAGE, transferred to Immobilon membrane, and immunoblotted with anti-RIIα, anti-RIIβ, and anti-actin mAbs. Compared with untreated cells, dexamethasone did not alter the total T cell content of either RII isoform or actin over time (data not shown). To test whether or not dexamethasone alters 8-Cl-cAMP-induced nuclear RIIβ translocation, normal T cells were cultured in the absence or presence of 10 nM dexamethasone for 18 h; the cells were then incubated for 1 h in the absence or presence of 20 μM 8-Cl-cAMP; the nuclei were separated; and, RIIβ was isolated by immunoprecipitation and quantified after immunoblotting with anti-RIIβ mAb. Compared with untreated cells, dexamethasone produced no significant change in cell viability. Moreover, dexamethasone did not alter 8-Cl-cAMP-induced nuclear translocation of RIIβ (Fig. 7). These results suggest that dexamethasone is unlikely to modify T cell RII subunit protein content or its subcellular localization.
The results of this work demonstrate that T lymphocytes from subjects with SLE may harbor a deficiency of PKA-II isozyme phosphotransferase activity. The prevalence of this isozyme deficiency in this study population was 37%. Our previous work has also revealed the presence of deficient PKA-I activity in the T cells of 80% of SLE subjects (13, 14, 15, 16). In this cohort of SLE subjects, one-third possessed a combined deficiency of both isozymes. On average, the mean specific activity of PKA-II was ∼60% that of age-, gender-, and racially matched controls. However, three subjects had profoundly low isozyme activities, less than 10% of the mean physiologic specific activity. By contrast, PKA-I isozyme activity in SLE T cells ranges between 20 and 25% of controls (16). Of particular interest is that there was no relationship between the extent of deficient PKA-II activity and SLE disease activity over a period of 4 years as quantified by a standardized disease activity index. These results mirror our recent findings for deficient PKA-I activity, in which diminished PKA-I activity was found to be persistent over time and independent of disease activity (14). Considering these data, it is reasonable to conclude that the profound decrement in total PKA activity reflects a deficiency of PKA-I and/or PKA-II isozyme activities in SLE T cells. We propose that a significant deficiency of PKA activity could markedly hinder effective signal transduction by impairing efficient substrate phosphorylation in SLE T cells.
In normal T cells, the PKA-II isozyme is predominantly found in the cytosol in its holoenzyme forms, RIIα2C2 and RIIβ2C2 (26). However, the amount of RIIβ protein is ∼4-fold higher than that of RIIα. Interestingly, T cells from SS disease controls actually express significantly increased amounts of cytosolic RIIβ compared with normal controls, yielding a markedly skewed RIIβ:RIIα ratio of 5.35:1. At present, however, the mechanism underlying these alterations in RII isoform expression in SS T cells remains uncertain. Although the RII isoforms are predominantly localized to the cytosol, a small amount of RIIβ isoform is constitutively present in the nucleus of normal primary T cells. The presence of constitutive nuclear RIIβ can be detected only by immunoblotting, for the amount is below the sensitivity of confocal immunofluorescence microscopy. This was true for T cells from both normal and SS disease controls. Activation of PKA-II by anti-CD3 + anti-CD28 + rIL-1α or 8-Cl-cAMP results in the separation of the RIIα and RIIβ subunits from the C subunit and the rapid translocation of RIIβ, but not RIIα, to the nucleus from the cytosol (36). Because this translocation enhances the amount of nuclear RIIβ, this process can be monitored by confocal immunofluorescence microscopy. RIIβ can first be detected in the nucleus by 30 min and peaks at 1 h. Interestingly, treatment of normal T cells with dexamethasone, a corticosteroid similar to that commonly used in the treatment of SLE, did not alter the subcellular localization of the RIIβ subunit or impede its translocation to the nucleus after activation of the RIIβ2C2 holoenzyme by 8-Cl-cAMP. Because both RII isoforms possess the nuclear localization sequence, KKRK, in their carboxyl-terminal regions (38), it is uncertain why RIIβ, but not RIIα, translocates to the nucleus after activation of PKA-II. At present, the role of RIIβ in the nucleus is being studied.
In SLE T cells, we observed an association between deficient PKA-II activity and enhanced translocation of RIIβ to the nucleus from the cytosol. On average, 60% of freshly isolated SLE T cells had identifiable nuclear RIIβ by confocal immunofluorescence microscopy. This compares with only 2–3% of normal and SS control T cells or 4% of SLE T cells that did not have deficient PKA-II activity. When the amount of RIIβ was quantified in isolated nuclear and cytosolic extracts, the content of nuclear RIIβ was increased by 54%, and that of cytosolic RIIβ was reduced by 60% compared with normal T cells. This shift produced a significant reduction in the ratio of cytosolic RIIβ:RIIα to 1.28:1 from 3.95:1 in normal T cells. The 6% difference in the amount of RIIβ between the cytosolic and nuclear compartments is probably insignificant, for it was within the error of the assay. However, it is pertinent to point out that, of the 21 SLE subjects that we analyzed, 29% had no detectable T cell cytosolic RIIβ protein. That both normal and SS control T cells have significantly higher cytosolic RIIβ protein content than do SLE T cells underscores the disease specificity of nuclear RIIβ translocation.
The association of a skewed RIIβ:RIIα ratio with diminished PKA-II activity in freshly isolated SLE T cells raised the possibility that nuclear RIIβ translocation could be triggered by the lupus microenvironment. SLE plasma often has increased levels of cytokines (e.g., IFN-α, IL-10) (39, 40), immune complexes (41), and complement fragments (42), which may bind to and potentially alter T cell signaling. However, an effect of the lupus microenvironment seems unlikely for two reasons. First, our previous experiments failed to demonstrate any effects of either IFN-α or immune complexes on PKA-catalyzed protein phosphorylation in normal T cells cultured in vitro. Contrariwise, culturing SLE T cells in vitro did not reverse the observed defect in PKA-catalyzed protein phosphorylation in SLE T cells (43). Second, in the present experiments, SLE T cells were propagated through 10 passages, and the progeny were analyzed. These T cell progeny had never been exposed to the lupus microenvironment and would, therefore, not be expected to exhibit any putative aberrant functions that freshly isolated T cells might express from exposure to that environment. Once these T cell progeny from established T cell lines reentered G0/G1 phase of the cell cycle, there was still no detectable cytosolic RIIβ; by contrast, control T cell progeny again expressed cytosolic RIIβ. Moreover, PKA-II activities remained markedly depressed and unchanged from that of freshly isolated SLE T cells whereas that of control T cells increased toward baseline activities of quiescent cells. These results are consistent with the idea that 1) PKA-II activity is diminished due to reduced/absent cytosolic RIIβ with which to form the holoenzyme, RIIβ2C2, and 2) RIIβ is retained in the nucleus.
Subcellular localization of PKA R and C subunits appears to be a principal mechanism to juxtapose the kinase to cAMP and its substrates. Longstanding evidence supports the concept that the PKA isozymes are localized to discrete regions of cells and that this is often cell-type specific (44, 45, 46, 47). In the human T cell, PKA-I associates with the plasma membrane fraction, whereas PKA-II is localized to the cytosol (26). Within the cytosol, the RIIα2C2 and RIIβ2C2 holoenzymes are compartmentalized to cytoskeletal elements and cytosolic organelles by attachment to anchoring structures termed A kinase anchor proteins (AKAPs) (48). RIIβ binds to AKAP75 in neuronal cells (49) and in T cells (Shook and G. M. Kammer, unpublished data), where it is likely to be in its holoenzyme form. Here, on its activation by cAMP, RIIβ2C2 holoenzyme dissociates to free RIIβ and C subunits and RIIβ can shuttle to the nucleus. This mechanism is shown in Fig. 8,A. After its release from the cAMP response element binding protein (CREB) heterodimer, our current evidence suggests that RIIβ is then exported from the nucleus to the cytosol where it can then bind AKAP75 and reform holoenzyme. However, the mechanism by which RIIβ is exported remains to be established. On the basis of our current data, we propose that in SLE T cells spontaneous activation of the RIIβ2C2 holoenzyme promotes aberrant nuclear RIIβ translocation and sequestration, resulting in deficient PKA-II activity (Fig. 8 B).
Two pertinent issues are currently under study. First, what is the stimulus that initiates RIIβ2C2 activation, resulting in RIIβ nuclear translocation? One mechanism may be cytokine receptor-mediated activation of the AC-cAMP-PKA pathway. To date, TGF-β is the only known cytokine that directly activates the AC-cAMP-PKA pathway in mesangial cells (50) and primary T cells (Choi and G. M. Kammer, unpublished data). However, synthesis of activated TGF-β by NK cells is markedly depressed in SLE (51), making it unlikely that the AC/cAMP/PKA pathway is being spontaneously activated via binding of TGF-β to its cell surface receptors. Moreover, PKA is unlikely to be activated by enhanced cAMP binding to RIIβ subunits, for we know that basal intracellular cAMP concentrations are comparable in unstimulated SLE and control T cells (52, 53). Additionally, if the PKA-II isozyme is like the PKA-I isozyme (15), the apparent Ka for PKA-II may be increased and would require higher concentrations of endogenous cAMP to activate the isozyme.
Second, why does the RIIβ subunit accumulate in the nucleus of SLE T cells? The RIIβ subunit has a nuclear localization signal (NLS), KKRK, in its carboxyl terminus. This sequence accounts for the capacity of RIIβ to enter the nucleus (38). However, the protein does not possess the consensus nuclear export signal (NES), XLXXXLXXLXLX (54). Instead, it has a partial sequence, VLDAMFEKLV (55), in which a hydrophobic phenylalanine (F) replaces the nonpolar leucine (L). This 10-aa stretch is positioned in the cAMP A-binding region between residues 165 and 174. Such replacements of one hydrophobic amino acid for another in NES sequences have been previously identified, as, for instance, in cyclin B1 (56). At present, however, it remains uncertain whether this partial sequence is a functional NES. If it is a functional NES, this may account for the nuclear-cytoplasmic shuttling observed in normal T cells as they reenter the G0/G1 phase of the cell cycle. Because RIIβ protein does not possess an apparent consensus nuclear retention signal, which could override the nuclear export signal resulting in retention of the protein in the nucleus (57), this mechanism cannot be invoked to account for the overexpression of nuclear RIIβ in SLE T cells. Currently, the mechanism of nuclear retention of RIIβ in SLE T cells remains to be established.
In summary, our results reveal that SLE T cells may harbor a deficiency of PKA-II isozyme activity that persists over time and is unassociated with disease activity. In about one-third of subjects, both PKA-I and PKA-II deficiencies can coexist. Of particular interest is the recognition that the mechanisms underlying these isozyme deficiencies are different. There is a significant reduction of RIβ > RIα protein that is associated with a marked reduction of RIβ > RIα transcripts in SLE T cells (16). Indeed, our current data suggest that there may be a pretranslational block of RIβ protein synthesis in the T cells of some SLE subjects (I. U. Khan and G. M. Kammer, unpublished data). By contrast, deficient RIIβ2C2 activity is a consequence of spontaneous activation of this holoenzyme, release of RIIβ, and its translocation to and retention in the nucleus. Long-term overexpression of nuclear RIIβ may alter transcriptional activation of genes, such as c-fos (Fig. 8 B).
This work was supported by grants from the National Institutes of Health (RO1 AR39501 to G.M.K. and RO1 AI42269 to G.C.T. and G.M.K.), the Lupus Foundation of America (to I.U.K. and G.M.K.), the General Clinical Research Center of the Wake Forest University School of Medicine (MO1 RR07122), the National Cancer Institute (5P30 CA12197), and the North Carolina Biotechnology Center (9510-1DG-1006). N.M. is a postdoctoral research fellow of the Arthritis Foundation.
Abbreviations used in this paper: SLE, systemic lupus erythematosus; SS, Sjögren’s syndrome; PKA, protein kinase A; PKA-I or -II, type I or II isozyme of PKA; RIIα/β, α or β isoform of regulatory subunit of PKA-II; RIIα/β2C2 holoenzyme, homodimer of RIIα or RIIβ isoform with C subunit; SLEDAI, SLE disease activity index; AC, adenylyl cyclase; PI, propidium iodide; ECL, enhanced chemiluminescence; ADU, arbitrary densitometric units; 8-Cl-cAMP, 8-chloro-cAMP; AKAP, A kinase anchor protein; CREB, cAMP response element binding protein; NES, nuclear export signal.