Expression of Constitutively Active cGMP-Dependent Protein Kinase Prevents Glucose Stimulation of Thrombospondin 1 Expression and TGF-β Activity

Monoclonal antibody 133 raised against human platelet TSP1 stripped of TGF-β activity was purified by our lab in a joint effort with the University of Alabama at Birmingham Hybridoma Core Facility (26). Stripped TSP1 (sTSP1) was purified from human platelets and depleted of TGF-β activity, as described previously, by gel permeation chromatography at alkaline pH (9). Anti-PKG-I was purchased from StressGen Biotechnology (Victoria, Canada). Rabbit polyclonal antiserum against rat fibronectin was purchased from Life Technologies (Gaithersburg, MD). Rabbit anti-collagen type IV antibody was purchased from Research Diagnostics (Flanders, NJ). Goat anti-mouse IgG and goat anti-rabbit IgG were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA). Monoclonal anti-TGF-β1, -β2, and -β3 antibody was purchased from R&D Systems (Minneapolis, MN). Luciferase assay reagent and passive lysis buffer were purchased from Promega (Madison, WI). Zeocin, Blasticidin, and Superscript one-step RT-PCR with platinum Taq were purchased from Invitrogen (Carlsbad, CA). Effectene transfection reagent was obtained from Qiagen (Valencia, CA).

RMCs were a gift from Dr. Anne Woods, University of Alabama at Birmingham. RMCs were cultured in RPMI-1640 media supplemented with 20% heat-activated fetal bovine serum (FBS), 5 mmol/l d-glucose, 2 mmol/l l-glutamine, 1% nonessential amino acids, 2 mmol/l sodium pyruvate, 10 μg/ml transferring, 5 ng/ml sodium selenite, and 0.6 international units/ml insulin. Serum-free RPMI-1640 media were used to quiescent RMCs. Transfection was carried out by using Effectene as described by the manufacturer. RMCs (50–70% confluent) were transfected with 0.8 μg pcDNA 6/TR (Invitrogen) for 18 h in RPMI-1640 media containing 20% FBS and selected with 3 μg/ml blasticidin. Once a stable cell line expressing the tetrocycline repressor (TetR) was obtained, these cells were cotransfected with 0.8 μg pcDNA4/TO/myc-His C/CD (a construct of the catalytic domain of G-kinase I α (bp 999-2016) or empty vector (pcDNA4/TO/myc-His C) as control cells. Stably transfected cells (labeled as RMC [tr/cd]) were selected with 3 μg/ml blasticidin and 500 μg/ml zeocin.

RMCs (tr/cd) were made quiescent in serum-free RPMI-1640 media with tetracycline (0.2–1.5 μg/ml) for 2 days and then were harvested in cold PEM buffer (20 mmol/l sodium phosphate [pH 6.8], 2 mmol/l EDTA,15 mmol/l β-mercaptoethanol, 0.15 mol/l NaCl, 0.1 mmol/l phenylmethysulfonyl fluoride, 10 μg/ml leupeptin, 10 μg/ml pepstatin A, and 5 μg/ml aprotinin) and homogenized (25). The extract was subjected to 10% SDS-PAGE and transferred to polyvinylidine fluoride membrane to detect catalytic domain of PKG by using polyclonal anti-PKG antibody.

Total RNA was extracted from RMCs and RMC (tr/cd). RNA (1 μg) was subjected to one-step RT-PCR (Invitrogen). Primers and conditions for amplification of the catalytic domain of PKG I genes were described previously (25).

To assess the biological activity of the expressed catalytic domain of PKG, PKG enzyme activity was measured in RMC (tr/cd) using methods described previously (24). PKG enzyme activity is expressed as nanomoles of peptide phosphorylated per minute per milligram of cell extract protein.

Cells were cultured and made quiescent for 48 h in the presence of 1 μg/ml tetracycline and then treated for the next 24 h with serum-free RPMI-1640 media containing either 5 or 30 mmol/l glucose. Conditioned media were harvested, and equal amounts of protein were subjected to SDS-PAGE gel (8%) under reducing conditions for detecting TSP1 and fibronectin. For detection of type IV collagen, cells were harvested in Laemmli sample buffer. After electrophoretic transfer to nitrocellulose membranes and blocking, the membranes were incubated with mouse monoclonal anti-TSP1 antibody 133 (1:5,000), rabbit anti-rat fibronectin antibody (1:1,000), or rabbit anti-type IV collagen antibody (1:5,000) for 1 h at room temperature. After washing, secondary antibody was used for the detection of immunoreactive bands with enhanced chemiluminescence detection system.

Cells were transiently transfected using Effectene with (0.3 μg) luciferase reporter plasmid containing the −2,033/+750 region of the human TSP1 gene promoter (a gift from Dr. Paul Bornstein, University of Washington). Luciferase activities were assayed 24 h posttransfection using the dual luciferase assay kit (Promega) according to the manufacturer’s directions.

Total and active TGF-β levels in the condition media were assayed using the plasminogen activator inhibitor-1 (PAI-1)/luciferase assay as described previously (12,27). The standard curve was obtained by using human recombinant TGF-β1 (R&D Systems), which was not affected by high glucose concentrations (30 mmo/l glucose).

RMC (tr/cd) were made quiescent in the absence or presence of tetracycline for 48 h and then treated with 30 mmol/l glucose for 24 h. After 24 h, media were removed and actinomycin D was added (designated as t = 0). After additional time points, cells were harvested. Northern analysis and hybridization for TSP1 and GAPDH were performed as described previously (12). Measurement of the ratio of TSP1/GAPDH at t = 0 (from actinomycin D treatment) was assigned a relative value of 100%.

Data are expressed as the mean ± SD. Statistical evaluation of the data was performed using the Student’s t test, considering a P value of <0.05 as significant.

We previously showed that high concentrations of glucose mediate increases in TSP1 expression and TSP1-dependent TGF-β bioactivity through down-modulation of NO-dependent PKG signaling (12). To directly investigate the role of PKG in regulation of TSP1 expression, we generated stable transfectants expressing the catalytic domain of PKG in RMCs (tr/cd) using an inducible expression system (tetracycline-induced gene expression [Tet-On]). Catalytic domain expression was determined in the cell lysate after 48 h in the presence of 0–1.5 μg/ml tetracycline (Fig. 1A). Induction of catalytic domain expression was time-dependent and sustained for at least 72 h (Fig. 1B). The RT-PCR results further confirmed the expression of a 1,000-bp band corresponding to the catalytic domain of PKG in RMCs (tr/cd) (Fig. 1C).

Activity of the catalytic domain was assessed using a selective PKG substrate, the BPDEtide (24). Cells were assayed in the presence or absence of cGMP to distinguish endogenous PKG activity from that of the cGMP-independent catalytic domain. As shown in Table 1, basal PKG activity was ∼4.5 nmol, and activity approximately doubled in the presence of cGMP. There was no difference in basal PKG activity between control cells and RMCs (tr/cd) without tetracycline induction. After 1 μg/ml tetracycline induction, PKG activity in RMCs (tr/cd) increased ∼2.8-fold compared with RMCs (tr/cd) without tetracycline induction. Moreover, addition of cGMP further increased PKG activity in RMCs (tr/cd) due to the activation of endogenous PKG by cGMP. These results indicate that the PKG catalytic domain expressed in RMCs (tr/cd) is fully active. This Tet-On system allows us to directly regulate the activity of PKG independent of cGMP levels and enables us to investigate the effects of PKG on glucose-mediated TSP1 gene expression, TGF-β bioactivity, and concomitant ECM expression.

Previously, we showed that high glucose inhibits NO production and intracellular cGMP levels. However, PKG protein levels were not altered (12). To determine the effects of high glucose on PKG activity, RMCs (tr/cd) were cultured in the presence of tetracycline (0–1.5 μg/ml) and then treated with high concentrations of glucose. Cell lysates were analyzed for PKG activity. As shown in Fig. 2A, high glucose concentrations significantly decreased endogenous PKG activity (tet = 0) (P < 0.001). The endogenous PKG activity in media with 30 mmol/l glucose is only 10% of that in 5 mmol/l glucose media. However, with the addition of cGMP (10 μmol/l), the endogenous PKG activity was essentially equivalent in cells exposed to either 5 or 30 mmol/l glucose, suggesting that the glucose-mediated decrease in intracellular cGMP levels is responsible for the reduced endogenous PKG activity (12). To the best of our knowledge, this is the first direct evidence that glucose downregulates PKG activity.

Furthermore, PKG activity was also measured upon tetracycline induction of RMCs (tr/cd). In the absence of added cGMP, 1 and 1.5 μg/ml tetracycline induced sufficient PKG activity to overcome glucose-mediated downregulation of PKG activity. Expression of the PKG catalytic domain and addition of cGMP to activate endogenous PKG resulted in similar levels of PKG activity under both 5 and 30 mmol/l glucose conditions (Fig. 2A). High concentrations of glucose did not change the expression of the catalytic domain compared with normal glucose (data not shown).

Consistent with these observations, glucose stimulation of TSP1 mRNA (Fig. 2B, 2C) and protein levels (Fig. 2D and E) decreased to basal levels (under 5 mmol/l glucose conditions) with tetracycline (1 and 1.5 μg/ml) induction. In addition, high glucose-mediated upregulation of TSP1 expression is TGF-β independent, since anti-TGF-β antibody could not alter TSP1 protein expression (Fig. 2F). Taken together, these data suggest that expression of the PKG catalytic domain to restore PKG activity independent of cGMP levels overcomes high glucose stimulation of TSP1 gene expression.

TSP1 has been shown to be a major physiological regulator of TGF-β activation. Previously, we have demonstrated that increased TSP1 expression is responsible for high glucose stimulation of TGF-β bioactivity (9–11). Furthermore, downregulation of NO-dependent PKG activation was responsible for glucose stimulation of TSP1 protein and TSP1-dependent TGF-β activation (12). Therefore we examined whether stimulation of PKG activity would block these increases in TGF-β activity. As shown in Fig. 3A, after tetracycline induction in RMCs (tr/cd) the elevated level of TGF-β bioactivity under high glucose conditions was reduced to basal levels. Moreover, exogenously added sTSP1 (2 or 10 nmol/l) overcame the actions of PKG in regulating TGF-β activity, suggesting that the effect of PKG on regulation of TGF-β activity occurs as a result of down regulation of TSP1 expression. Total TGF-β levels were not affected by either increased PKG activity or exogenously added sTSP1 (Fig. 3B). In the absence of tetracycline, treatment of the cells with sTSP1 resulted in a similar fold increase in active TGF-β levels under either normal or high glucose conditions. No significant changes were observed in the levels of total TGF-β with various doses of sTSP1 treatment (data not shown).

The specificity of the luciferase response for TGF-β signaling in PAI-1/luciferse assay was previously tested by using an anti-TGF-β antibody. The antibody to TGF-β nearly completely blocked the luciferase reporter activity in those assays. In addition, results were also confirmed by using the R&D systems Quantikine enzyme-linked immunosorbent assay for TGF- β1 (12).

TGF-β has been shown to stimulate ECM production under high glucose conditions, which play an important role in the development of diabetic nephropathy (28–29). Our lab has demonstrated that high glucose stimulation of ECM protein expression in mesangial cells is dependent on TSP1-mediated activation of TGF-β (11). Without tetracycline induction, high glucose stimulated fibronectin and type IV collagen production in RMCs (tr/cd). These effects of glucose are TGF-β dependent because fibronectin and type IV collagen upregulation by high glucose could be blocked by anti-TGF-β neutralizing antibody. In RMCs (tr/cd) with tetracycline induction, increased PKG activity inhibited high glucose-mediated increases in fibronectin and type IV collagen protein levels (Fig. 4).

Our data show that PKG regulates both TSP1 mRNA and protein levels. To better understand how PKG regulates transcription, we examined steady-state TSP1 mRNA levels. As shown in Fig. 5A and B, in control cells high glucose increased TSP1 mRNA levels in a time-dependent manner with a fourfold increase at 24 h as compared with normal glucose. In RMCs (tr/cd) without tetracycline induction, high glucose similarly increased TSP1 mRNA levels (Fig. 5C and D). However, with tetracycline induction of PKG activity, glucose stimulation failed to upregulate TSP1 mRNA.

TSP1 mRNA stability was unaffected with expression of the catalytic domain (Fig. 6). Consistent with the report of Donoviel et al. (30), TSP1 mRNA was degraded with a half-life of ∼5.5 h. Our data suggest that increased PKG activity does not alter the rate of TSP1 mRNA turnover.

Alternately, PKG might affect the transcriptional activity of the TSP1 promoter. The TSP1 gene promoter region has been characterized (31–32). With the transfection of a full-length human TSP1 gene promoter-reporter construct (−2,033 to +750) into RMCs (tr/cd), we found that increased PKG activity inhibited high glucose-stimulated reporter activity (Fig. 7). Anti-TGF-β antibody had no effect on reporter activity, suggesting that the effect of PKG on TSP1 transcription is TGF-β independent, consistent with our previous observations (11,12). Furthermore, region −1,172 to −548 of the human TSP1 promoter, containing binding sites for transcription factors such as AP-1, Sp-1, SRF, and upstream stimulatory factor (USF), is involved in transcriptional repression by PKG (data not shown). Recently, Volpert et al. (33) showed that Id1 represses TSP1 promoter activity in mouse embryo fibroblasts, and a strong repressive region in −1,120 to −1,310 of the murine TSP1 promoter was identified. This transcriptional repression region overlaps with part of the PKG repressive region in the TSP1 promoter, suggesting that Id might be involved in the mechanisms by which PKG represses TSP1 gene transcription. Id, expressed in mesangial cells (34), is a family of dominant-negative helix-loop-helix (HLH) proteins that block cell-specific transcription mediated by basic HLH transcription factors (36). USF, a member of the basic HLH leucine zipper family (37), has been shown to be a glucose-responsive transcription factor (38). Glucose-induced expression of Id1 (35) might bind to USF, leading to PKG-mediated repression of TSP1 gene transcription under high glucose conditions. Such a mechanism would be consistent with the lack of PKG regulation of basal TSP1 gene transcription observed in previous studies (39) in which basal TSP1 mRNA levels only slightly decreased in aortic smooth muscle cells 48 h after transfection with the constitutively active catalytic domain of PKG-I. Under basal conditions, posttranscriptional mechanisms appear to be involved. The roles of PKG in modulating transcription factors that regulate TSP1 promoter are currently under investigation.

In summary, our data show that increased PKG activity abolished glucose stimulation of TSP1-dependent TGF-β bioactivity, resulting in decreased ECM protein levels. These data suggest that PKG-regulated TSP1-dependent TGF-β activity plays a role in the pathogenesis of diabetic nephropathy. Gene transfer of the catalytic domain of PKG to modulate PKG activity might provide new strategies for the treatment of diabetic renal fibrosis.

This work was supported by National Institutes of Health (NIH) Grant R01-DK54624 to J.M.U., NIH HL53426 to T.M.L., and a Juvenile Diabetes Research Foundation postdoctoral fellowship (3-2002-345) to S.W.

We thank Dr. Paul Bornstein, University of Washington, for the gift of TSP1 gene promoter reporter constructs.