Nuclear factor (NF)-κB and signal transducer and activator of transcription 3 (STAT3) play a critical role in diabetic nephropathy (DN). Sirtuin-1 (SIRT1) regulates transcriptional activation of target genes through protein deacetylation. Here, we determined the roles of Sirt1 and the effect of NF-κB (p65) and STAT3 acetylation in DN. We found that acetylation of p65 and STAT3 was increased in both mouse and human diabetic kidneys. In human podocytes, advanced glycation end products (AGEs) induced p65 and STAT3 acetylation and overexpression of acetylation-incompetent mutants of p65 and STAT3 abrogated AGE-induced expression of NF-κB and STAT3 target genes. Inhibition of AGE formation in db/db mice by pyridoxamine treatment attenuated proteinuria and podocyte injury, restored SIRT1 expression, and reduced p65 and STAT3 acetylation. Diabetic db/db mice with conditional deletion of SIRT1 in podocytes developed more proteinuria, kidney injury, and acetylation of p65 and STAT3 compared with db/db mice without SIRT1 deletion. Treatment of db/db mice with a bromodomain and extraterminal (BET)-specific bromodomain inhibitor (MS417) which blocks acetylation-mediated association of p65 and STAT3 with BET proteins, attenuated proteinuria, and kidney injury. Our findings strongly support a critical role for p65 and STAT3 acetylation in DN. Targeting protein acetylation could be a potential new therapy for DN.
Diabetic nephropathy (DN) is the most common cause of end-stage renal disease (ESRD) in the U.S., and its incidence has been increasing worldwide over the past decade (1). Current interventions provide only partial therapeutic effects, and therefore it is critical to develop more effective therapy for DN.
Many studies have suggested a role for advanced glycation end products (AGEs) in the pathogenesis of DN. Nondiabetic animals infused with AGEs develop histological changes similar to those found in animals with DN (2,3). Reduction of AGE accumulation by lowering dietary AGE consumption or by pharmacologic inhibition of AGE formation has been shown to ameliorate microvascular diseases and DN in animal models (2,4,5). We have demonstrated previously that both circulating and matrix-bound AGEs cause podocyte apoptosis through activation of the receptor for AGE (6).
Multiple transcription factors (TFs) that are activated under the diabetic condition also are known to mediate hyperglycemia- and AGE-induced pathologic changes in DN. Systems-based analyses of gene expression patterns in the diabetic kidney suggest that Janus kinase–signal transducer and activator of transcription (JAK/STAT) and nuclear factor (NF)-κB signaling are major pathways activated in the kidneys of humans with DN (7,8). We demonstrated previously that FOXO4 is a major TF that mediates AGE-induced podocyte apoptosis by activating Bim-1, a proapoptotic protein (6,9). TF activation is typically determined by its phosphorylation status. However, recent evidence suggests that protein acetylation also is required for TF activation. Sirtuin-1 (SIRT1), an enzyme that mediates NAD+-dependent deacetylation of target substrates, regulates the activity of many TFs by targeting them for deacetylation. SIRT1 regulates the acetylation status of the FOXO family of TFs, thereby affecting their downstream gene expression (10,11). We showed that SIRT1 inhibits podocyte apoptosis by deacetylating FOXO4 (6,9). Several studies suggest that the transcriptional activity of STAT3 also is negatively regulated by SIRT1 (12–15). SIRT1 mediates anti-inflammatory effects by inhibiting NF-κB transcriptional activity via its acetylation-mediated association with bromodomain and extraterminal (BET) bromodomain proteins such as BRD4 (16–18). We found that SIRT1 expression is suppressed by AGE in cultured podocytes and in human diabetic kidneys (9).
Based on these findings, here we attempt to address 1) the acetylation status of TFs in the diabetic kidney; 2) the effect of AGE inhibition in diabetic db/db mice on SIRT1 expression, TF acetylation, and podocyte apoptosis; 3) the role of podocyte SIRT1 in the progression of DN and regulation of TF acetylation; and 4) the therapeutic effects of bromodomain inhibitors in DN.
Research Design and Methods
Female db/db and db/m mice (6 weeks old) in C57BLKS background were obtained from The Jackson Laboratory (Bar Harbor, ME). When mice developed hyperglycemia (8 weeks of age), they were randomized to receive either pyridoxamine (PYR; n = 6; 250 mg/kg/day in drinking water; Biostratum Inc., Durham, NC) or vehicle (n = 6). Mice were housed individually, and the concentration of PYR in drinking water was adjusted daily based on the animal’s water intake. Floxed SIRT1 mice (SIRT1fl/fl) were obtained from The Jackson Laboratory. These mice were crossed with podocin (Pod)-Cre+/− mice to generate podocyte-specific SIRT1 knockout mice (Pod-Cre;SIRT1fl/fl). Pod-Cre+/−;SIRT1fl/fl mice were further crossed with db/db mice to generate Pod-Cre+/−;SIRT1fl/fl;db/m mice, which were further crossed with each other to generate Pod-Cre+/−;SIRT1fl/fl;db/db mice. Pod-Cre+/−;db/db and Pod-Cre+/−;db/m mice also were generated. Therefore, 4 groups of mice were included in the study: Pod-Cre+/−;SIRT1fl/fl;db/db, Pod-Cre+/−;SIRT1fl/fl;db/m, Pod-Cre+/−;db/db, and Pod-Cre+/−;db/m.
Diabetic db/db and nondiabetic db/m mice were fed either a control vehicle (DMSO) or MS417 by daily gavage at a concentration of 0.08 mg/kg, as described previously (19). The mice (n = 6) were fed with this compound daily from the age of 8 weeks to 20 weeks. Unrestricted food and water were provided throughout the duration of the experiment. The mice were killed at 20 weeks of age.
For all experiments, glycemia and glycosuria were monitored in diabetic animals twice per week. After the mice were killed, blood, urine, and kidney tissue were collected. Iron beads were perfused in one kidney for glomerular isolation; the other kidney was perfused with 4% paraformaldehyde for histology and immunostaining. All animal studies were performed according to the protocols approved by the Institutional Animal Care and Use Committee at the Mount Sinai School of Medicine.
Quantification of Urine Albumin
Urine protein was quantified by ELISA for albumin with a kit from Bethyl Laboratories Inc. (Houston, TX). Urine creatinine levels were measured in the same samples. The urine albumin excretion rate is expressed as the ratio of albumin to creatinine.
Histology and Morphological Analysis
Kidney histology was examined after periodic acid Schiff staining. The glomerular volume and mesangial area were determined by examining periodic acid Schiff–stained sections using a digitizing tablet and video camera, as previously described (5). The relative mesangial area was expressed as the mesangial-to-glomerular surface area (percentage).
Isolation of Glomeruli
Paraffin-embedded sections of mouse and human kidneys were incubated with primary antibodies and then with biotinylated second antibodies, followed by incubation with an avidin–biotin–peroxidase complex; they were developed using substrate provided by Vector Laboratories. The human biopsy and nephrectomy samples were obtained under a protocol approved by the institutional review board of Icahn School of Medicine at Mount Sinai (HS#: 11–02130). The following antibodies were used in the study: SIRT1 (cat. no. 07131; Millipore), acetyl-STAT3 (cat. no. 2523; Cell Signaling Technology), phosphor-STAT3 (cat. no. 9145; Cell Signaling Technology), acetyl-p65 (cat. no. 3045; Cell Signaling Technology), phosphor-p65 (cat. no. 3033; Cell Signaling Technology), and synaptopodin (provided courtesy of Dr. Peter Mundel). The dilutions of both acetyl-p65 and acetyl-STAT3 antibodies for immunostaining were 1:50. Negative controls for acetyl-STAT3 and acetyl-p65 were obtained by preabsorption with a synthetic acetylated peptide corresponding to residues surrounding Lys310 of p65 or Lys685 of STAT3.
Quantification of Immunostaining
ImageJ 1.26t software was used to measure the level of immunostaining in the glomeruli. First, the images were converted to 8-bit grayscale. Next, the glomerular region was selected for measurement of area and integrated density. Then, the background intensity was measured by selecting three distinct areas in the background with no staining. The corrected optical density was determined as shown in the equation: corrected optical density = ID − (A × MGV), where ID is the integrated density of the selected glomerular region, A is the area of the selected glomerular region, and MGV is the mean gray value of the background readings (21). Synaptopodin staining was quantified by counting the percentage of positive staining area per glomerulus. The Sirt1 staining was quantified by counting the number of positively stained cells per glomerulus.
Conditionally immortalized human podocytes (courtesy of Dr. Moin Saleem, Bristol, U.K.) were cultured as described previously (22). Endotoxin-free AGE-BSA was prepared from endotoxin-free, lyophilized BSA (fraction IV; Sigma) and glucose, as described previously (6). AGE levels were measured by ELISA, as described previously by Makita et al. (23).
Human STAT3 cDNA plasmid was purchased from Open Biosystems (clone ID 3347434). The K685R mutation was generated following the Stratagene Site-Directed Mutagenesis protocol. The mutant primer sequence is 5′-CTC TGG CCG ACA ATA CCT TCC GAA TGC CTC CTC-3′. The wild-type and mutant full-length STAT3 sequences were amplified using a forward primer with a Bam1H cutting site (5′-TTG GAT CCG CCA CCA TGG CCC AAT GGA ATC AGC TAC-3′) and a reverse primer with an Xma1 cutting site (5′-CCG GCC CGG GTC ACA TGG GGG AGG TAG CGC-3′). PCR product then was cloned, first into a T-vector (Promega A1360) and finally into a pHR-CMV-MCS-IRES-gfp-delta B vector using Bam1H and Xma1. Human wild-type and K310R mutant p65 plasmids were also generated. PCR product was obtained with the following primers: 5′-CTA GAC TAG TTT AGG TGC TGA TCT GAC TCA-3′ (sense) and 5′-TAG ATA TCT TAG GAG CTG ATC TGA CTC AGC-3′ (antisense). The strategy of cloning into pHR-CMV-MCS-IRES-gfp-delta B vector was the same as that of STAT3, except endonucleases were used for Bam1H (5′ end) and Sma1 (3′ end).
Using the PolyJet transfection method, 50% confluent 293T cells were transfected with the following three plasmids: 1 μg pVSV.G, 2 μg pPACK together with 3 μg gene of interest plasmid of either empty pHR-CMV-MCS-IRES-gfp-delta B vector, or wild-type STAT3, or K685R STAT3, or wild-type p65, or K310R p65 plasmid. Medium containing viral particles was centrifuged at 5,000 rpm for 4 h at 4°C. The pellet was dissolved in 1 mL serum-free RPMI medium.
TUNEL fluorescent staining (Millipore) was used to determine the apoptosis rate in kidney sections of mice. We counted all glomeruli with at least one apoptotic podocyte, and the percentage of glomeruli with apoptotic cells was shown as described previously (24). At least 100 glomeruli per mouse were examined.
Quantitative reverse-transcriptase PCR was performed using Sybr Green Master Mix (Applied Biosystems). The 2-ΔΔCp method was used for the analysis of relative gene expression. Primers were designed using Primer-Blast (National Center for Biotechnology Information, Bethesda, MD) to span at least one intron; the sequences of these primers are listed in Supplementary Table 1. Gene expression was normalized to GAPDH as a housekeeping gene.
Nuclear proteins were extracted from glomeruli in a buffer containing a protease inhibitor cocktail and tyrosine and serine-threonine phosphorylation inhibitors. Immunoblotting was performed using the same specific antibodies as described above for immunostaining. The antibody for β-actin was from Sigma (cat. no. a2228). Densitometry analysis for quantification of Western blots was performed as described previously (25).
Chromatin Immunoprecipitation Assay
Human podocytes were infected with the following viruses: green fluorescent protein, wild-type and K658R STAT3 or green fluorescent protein, wild-type and K310R p65 (1:100 V/V) for 3 days and then treated with either AGE-BSA (100 μg/mL) or BSA (100 μg/mL) for an additional 24 h. Chromatin immunoprecipitation assay was performed according to the EZ-CHIP protocol (Millipore catalog no. 17–371) (26). Immunoprecipitated DNA samples were subjected to PCR analysis, normalized to input, and expressed as a percentage of input.
Reporter Gene Assay
Podocytes were transfected with luciferase reporter vectors for NF-κB or STAT3, as described by Zhang et al. (19). Luciferase activity was measured using a luminometer, and the data were normalized for the amount of protein.
Data are expressed as mean ± standard deviation. The unpaired t test was used to analyze data between two groups. ANOVA followed by Bonferroni correction was used when more than two groups were present. All experiments were repeated at least three times, and representative experiments are shown. P values <0.05 were considered statistically significant.
Acetylation of TFs Is Increased in Diabetic Kidneys
We previously found that acetylation of FOXO4 was significantly increased in the kidney of diabetic db/db mice compared with nondiabetic db/m mice (9). Here, we demonstrated that staining of both acetyl-p65 and acetyl-STAT3 was significantly higher in glomeruli of diabetic db/db mice compared with those of nondiabetic db/m mice (Fig. 1A and B). Western blots of isolated glomeruli from these mice revealed a higher level of phosphorylation and acetylation for p65 and STAT3 in diabetic db/db mice compared with db/m mice (Fig. 1C and D and Supplementary Fig. 1). Immunostaining of kidney biopsy samples revealed that acetylation of p65 and STAT3 was increased in both glomerular and tubular compartments of kidneys of patients with DN compared with patients with minimal change disease and normal kidney tissues from nephrectomy patients (Fig. 1E and F). The clinical characteristics of these patients are summarized in Supplementary Table 2.
NF-κB and STAT3 Acetylation Is Required for Transcriptional Activation of Target Genes
Since we have previously shown that AGE suppresses SIRT1 expression in podocytes (9), here we tested whether AGE increases the acetylation of these TFs in podocytes. Using Western blot analysis, we found that AGE treatment induced both phosphorylation and acetylation of p65 and STAT3 in podocytes compared with BSA treatment (Fig. 2A and B). These data, together with our previous findings, suggest that AGE likely induces the acetylation of p65 and STAT3 by inhibiting SIRT1 expression.
To confirm that acetylation of p65 and STAT3 is required for the transcriptional activation of their target genes, we generated mutants of p65 (K310R) and STAT3 (K685R), in which the major acetylated lysine residues were mutated. First, we confirmed the overexpression of these mutants in podocytes by Western blot analysis (Fig. 2C–F). We confirmed that overexpression of these mutants in podocytes inhibited tumor necrosis factor-α–induced NF-κB activation and interleukin-6–induced STAT3 activation as measured by luciferase reporter assays (Fig. 2G and H). Then, we found that overexpression of these mutants in podocytes suppressed AGE-induced expression of NF-κB and STAT3 target genes as measured by real-time PCR (Fig. 3A–D), thus confirming that acetylation of NF-κB and STAT3 is required for their transcriptional activation in podocytes. We selected CXCL5 and CCL17 as NF-κB target genes and SOC3 and Casp9 as STAT3 target genes based on previous studies (9).
Next, we determined whether acetylation of p65 and STAT3 is required for protein-DNA binding of TFs to the promoter of their target genes. Using chromatin immunoprecipitation assay, we found that AGE-induced binding of p65 and STAT3 in the promoter sites of their target genes were markedly inhibited by the overexpression of p65 (K310R) and STAT3 (K685R) in podocytes (Fig. 3E–H). These data further support the critical role of acetylation in the transcriptional regulation of NF-κB and STAT3 target genes in AGE-treated podocytes.
PYR Restores Sirt1 Expression and Reduces TF Acetylation in Diabetic db/db Mice
Next we determined whether the inhibition of AGE formation in db/db diabetic mice using a specific AGE inhibitor, PYR, restores Sirt1 expression, reduces TF acetylation, and attenuates podocyte apoptosis and proteinuria in diabetic db/db mice. In Supplementary Table 3, we summarize the body weight, kidney weight, and blood glucose concentrations of db/m and db/db mice treated with either vehicle or PYR. We found that the kidney weights of db/db mice treated with vehicle were higher than db/m mice treated with vehicle, but kidney hypertrophy was prevented in db/db mice treated with PYR. Diabetic db/db mice treated with vehicle developed significant proteinuria, glomerular hypertrophy, mesangial expansion, and podocyte loss compared with nondiabetic db/m mice (Fig. 4A–G). However, db/db mice treated with PYR developed less proteinuria, glomerular hypertrophy, mesangial expansion, and podocyte loss than db/db mice treated with vehicle (Fig. 4A–G). SIRT1 expression was partially restored in the glomeruli of db/db mice treated with PYR but not in db/db mice treated with vehicle (Fig. 5A–D). Consistent with previously findings, synaptopodin expression also was reduced in diabetic kidneys but was restored by PYR treatment (Fig. 5A–D). By immunostaining, we found that the acetylation status of both p65 and STAT3 in the glomeruli of db/db mice treated with PYR was lower than that in db/db mice treated with vehicle (Fig. 5E–H). Taken together, these data suggest that renal protection afforded by PYR is associated with the restoration of SIRT1 expression and inhibition of TF acetylation. However, whether suppression of SIRT1 and increased TF acetylation contribute to DN remains to be determined.
Mice With Podocyte-Specific Sirt1 Knockout Are More Susceptible to DN
To determine the role of SIRT1 in podocyte injury and the development of DN, we generated podocyte-specific SIRT1 knockout mice (Pod-Cre+/−;SIRT1fl/fl) by crossing Pod-Cre+/− mice with mice that are homozygous for a floxed SIRT1 allele (SIRT1fl/fl). Pod-Cre+/−;SIRT1fl/fl mice then were bred with db/db mice to generate Pod-Cre+/−;SIRT1fl/fl;db/m and Pod-Cre+/−;SIRT1fl/fl;db/db mice, which were used in subsequent experiments. Pod-Cre+/−;db/db and Pod-Cre+/−;db/m mice were used in the study as the control mice. The body weight, kidney weight, and blood glucose concentrations at the time the mice were killed are summarized in Supplementary Table 4. The kidney weight of Pod-Cre+/−;db/db mice was higher than that of Pod-Cre+/−;db/m mice, and a further increase was observed in Pod-Cre+/−;SIRT1fl/fl;db/db mice compared with Pod-Cre+/−;db/db mice. By immunostaining, we confirmed that SIRT1 expression was low in Pod-Cre+/−;db/db mice and even lower in Pod-Cre+/−;SIRT1fl/fl;db/m and Pod-Cre+/−;SIRT1fl/fl;db/db mice when they were compared with Pod-Cre+/−;db/m mice (Supplementary Fig. 2A–C). Immunostaining of synaptopodin in Pod-Cre+/−;db/db and Pod-Cre+/−;SIRT1fl/fl;db/db mice was markedly less than that in the corresponding nondiabetic controls (Supplementary Fig. 2A–C). We found that Pod-Cre+/−;db/db mice developed more proteinuria, mesangial expansion, and podocyte loss (Fig. 6A–D) than nondiabetic control mice. However, Pod-Cre+/−;SIRT1fl/fl;db/db mice developed more severe proteinuria, mesangial expansion, and podocyte loss than Pod-Cre+/−;db/db mice (Fig. 6A–D). These findings suggest that podocyte-specific knockout of SIRT1 aggravates podocyte injury and proteinuria in db/db mice. Next we determined the acetylation status of p65 and STAT3 in glomeruli of these mice. By immunostaining, we found that acetylation of p65 and STAT3 was significantly increased in most glomerular cells of Pod-Cre+/−;db/db mice compared with nondiabetic control mice (Supplementary Fig. 3A–D). However, staining for p65 and STAT3 acetylation seemed to be more pronounced in the podocytes of Pod-Cre+/−;SIRT1fl/fl;db/db mice compared with those of Pod-Cre+/−;db/db mice. These data suggest that deletion of SIRT1 specifically in podocytes promotes acetylation of p65 and STAT3, leading to more severe DN.
Effects of Bromodomain Inhibitors on the Development of DN
The above data suggest that acetylation of p65 and STAT3 contributes to the development and progression of DN. Although SIRT1 is known to modulate the acetylation status of these TFs, recent studies suggest that existing SIRT1 agonists may not be specific for SIRT1 (27,28). In addition, SIRT1 agonists may have both protective and harmful effects (29). Therefore, we have developed another approach to modulate TF acetylation as a potential therapy for DN. The bromodomain, which is the conserved structural module present in transcription-associated proteins and histone acetyltransferases, functions as the acetyl-lysine binding domain that is responsible for almost all acetylation-mediated protein-protein interactions in gene transcriptional activation in chromatin (30,31). Our recent study showed that blocking BRD4 association with K310-acetylated p65 using our recently developed small-molecule BET-specific bromodomain inhibitor (MS417) effectively downregulates NF-κB–mediated inflammatory response in kidney cells through inhibition of NF-κB–directed transcriptional activation of cytokines and attenuates kidney injury in HIV-1 transgenic mice (19). The specificity of MS417 on the inhibition of BRD4 binding with acetyl-residue (K310) of p65 also was confirmed in the study (19).
Here, we further confirmed that MS417 also attenuated AGE-induced p65 acetylation and phosphorylation in podocytes (Fig. 7A and B). Since BRD4 may also interact with acetyl-lysine residues of STAT3, we tested whether MS417 could inhibit STAT3 acetylation. We found that MS417 also inhibited AGE-induced STAT3 acetylation and phosphorylation in podocytes (Fig. 7C and D). Why the inhibition of acetylation of p65 and STAT3 by MS417 also reduced the phosphorylation of these TFs remains unclear. We speculate that inhibition of p65 and STAT3 acetylation may facilitate dephosphorylation of p65 and STAT3. However, this will need additional confirmation in future studies.
Next we determined whether MS417 improves kidney injury in diabetic db/db mice. We treated both male db/db and db/m mice with either vehicle or MS417 for a total of 12 weeks. Kidney hypertrophy observed in db/db mice was attenuated by MS417 treatment (Supplementary Table 5). MS417 treatment significantly reduced proteinuria and mesangial expansion in diabetic db/db mice compared with the mice treated with vehicle (Fig. 7E–G). The corresponding target genes of NF-κB and STAT3 that increased in the glomeruli of db/db mice also were suppressed by MS417 treatment (Supplementary Fig. 4A–D). These data suggest that small-molecule inhibition of the BET bromodomains is a potential novel therapeutic strategy for DN.
This is the first study to delineate the role of acetylation of NF-κB and STAT3 as a molecular mechanism of DN. We hypothesized that in diabetic kidneys, TF acetylation is increased because SIRT1 expression is inhibited by AGEs. We confirmed that modulation of TF acetylation—either by inhibition of AGE formation or by directly blocking the interaction between lysine-acetylated TFs and bromodomain-containing proteins—mitigated the development of DN, whereas knockout of SIRT1 in the podocytes aggravated DN (summarized in Fig. 8). Our results strongly support a critical role for downregulation of SIRT1 and hyperacetylation of STAT3 and NF-κB in the pathogenesis of DN. In addition, our studies suggest that several potential interventions could be developed to reduce TF acetylation as potential therapies for DN (Fig. 8).
We found that lysine acetylation is required for transcriptional activation of NF-κB and STAT3. This is consistent with previous studies showing that acetylation of NF-κB is required for transcriptional activation of its target genes (16,32). Previous unbiased analyses reveal that many target genes of NF-κB and STAT3 are activated in DN (7,8). The role of NF-κB and STAT3 in DN also is supported by other studies (33–36). Acetylation of other TFs may also play a role in DN. We have shown a critical role for FOXO4 acetylation in podocyte injury in DN (6). Resveratrol has been shown to reverse acetylation of Smad3 to inhibit transforming growth factor-β1–induced upregulation of collagen IV and fibronectin (37), and Smad3 plays a key role in DN (38). A large body of evidence suggests that p53 mediates the apoptosis of podocytes and tubular epithelial cells in DN (39,40). SIRT1 has been shown to regulate p53 activity by targeting it for deacetylation (41–43). These studies suggest that acetylation of TFs could be a key event leading to DN.
Many studies suggest that SIRT1 affects diseases of aging in mammals, such as diabetes, cancer, and inflammation (44). The role of SIRT1 also has been studied in kidney diseases, including DN (45,46). Hasegawa et al. (47) recently reported that mice with proximal tubular (PT)-specific SIRT1 knockout developed albuminuria, whereas overexpression of SIRT1 in PT protected mice from diabetic kidney injury. They demonstrated that the deficiency of SIRT1 in PT upregulates the expression of claudin-1 in podocytes, which causes podocyte dysfunction, glomerular filtration barrier disruption, and albuminuria in mice. Our mice with podocyte-specific Sirt1 knockout did not develop albuminuria in the basal condition; however, the mice manifested more severe kidney injury when they were diabetic. The discrepancy between these two models suggests that SIRT1 may serve different roles in podocytes and PT. Further studies are required to delineate the cellular mechanism that underlies the phenotypic differences between these two SIRT1 loss-of-function models.
The beneficial effects of SIRT1 likely occur through deacetylation of TFs. Since there are currently no proven specific SIRT1 agonists (27,28), the development of other approaches to modulate TF acetylation is important. Development of small molecules as epigenetic tools to block protein–protein interactions provided by bromodomains is a new therapeutic approach (48). Several cell-permeable small molecules that bind competitively to acetyl-lysine recognition motifs or bromodomains have been developed recently as cancer therapy (49–51). BRD4 belongs to the BET protein family that contains two bromodomains in tandem. Recent studies show that BET proteins play an important role in coordinating gene transcriptional activation in an acetylation-sensitive manner (52). In our previous studies we showed that our BET-specific bromodomain inhibitor (MS417) effectively attenuates NF-κB transcriptional activation of proinflammatory genes in kidney cells and ameliorates inflammation and kidney injury in HIV-1 transgenic mice (Tg26) (19). NF-κB–mediated inflammation is also a major pathology in DN (7), and our current data suggest that MS417 also inhibited AGE-induced NF-κB acetylation and activation in human podocytes. In addition, MS417 also inhibited STAT3 acetylation, another major pathway involved in DN (8). We confirmed here that MS417 reduces proteinuria and podocyte injury in diabetic db/db mice. In addition, because both NF-κB and STAT3 are key regulators of inflammation and because inflammation is a major contributor to renal fibrosis, we believe that MS417 also will attenuate renal fibrosis in DN. Consistent with this, we demonstrated previously that MS417 treatment improves renal fibrosis in Tg26 mice (19). Taken together, these data suggest that MS417 or its analogs could be developed as a potential new therapy to treat patients with DN.
In summary, our studies suggest that acetylation of NF-κB and STAT3 is increased in diabetic kidneys, contributing to the pathogenesis of DN likely through regulation of the expression of their target genes. Inhibition of AGE restores SIRT1 expression and reduces acetylation of STAT3 and NF-κB, and this is likely one of the mechanisms that AGE inhibitors use to attenuate kidney injury in diabetes. Knockout of SIRT1 in podocytes aggravates diabetes-induced kidney injury, probably by promoting TF acetylation in podocytes. Finally, we found that BET bromodomain inhibitors, which directly target acetyl-lysine residues of TFs, could be developed as a novel therapeutic approach for treatment of DN. Therefore, our studies provide not only the novel mechanism of but also a potential new therapy for DN.
See accompanying article, p. 2432.
R.L. and Y.Z. contributed equally to this work.
Funding. J.C.H. is supported by VA Merit Award (1I01BX000345), 973 fund (2012CB517601), and National Institutes of Health (NIH) grants 1R01-DK-078897 and 1R01-DK-088541. P.Y.C. is supported by NIH grant 5K08-DK-082760. Y.Z. is supported by the National Natural Science Foundation of China for Young Investigators (1999-30901944) and the Shanghai Bureau of Health for Young Investigators (2011-XYQ2011059). M.-M.Z. is supported by NIH grants R01-HG-004508 and R01-CA-87658.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. R.L., Y.Z., X.L., and H.C. performed the experiments and analyzed the data. B.J. provided human kidney biopsy samples and edited the manuscript. M.-M.Z. provided MS417 and edited the manuscript. P.Y.C. and J.C.H. designed the experiments and wrote the manuscript. J.C.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.